Build branch main with version main (f52978a)

Build pipeline: viash-hub.rnaseq.main-k4ttq

Source commit: f52978a0e2

Source message: Fix summarizedexperiment build (#42)
This commit is contained in:
CI
2025-05-06 20:05:48 +00:00
parent d08ba5f5fa
commit c00bcda3f8
275 changed files with 31239 additions and 8651 deletions

View File

@@ -1,9 +1,10 @@
# RNAseq version x.y.z
# rnaseq v0.2.1
## Changes
## Bug fixes
- Add resources labels
- Increase the number of components that are fetched from BioBox
* Fix `summarizedexperiment` build PR (#42).
* Fix an issue with the `deseq2_qc` component not being able to create the DESeq2 object (PR #41).
## Known issues

View File

@@ -1,13 +1,13 @@
name: rnaseq
viash_version: 0.9.0
viash_version: 0.9.2
source: src
target: target
info:
test_resources:
- path: gs://viash-hub-test-data/rnaseq/v1
- path: gs://viash-hub-resources/rnaseq/v1
dest: testData
config_mods: |

View File

@@ -0,0 +1 @@
Optional!

View File

@@ -0,0 +1 @@
Required!

View File

@@ -68,8 +68,16 @@ test_resources:
engines:
- type: docker
image: rocker/r2u:22.04
image: debian:latest
setup:
- type: apt
packages:
- libcurl4-openssl-dev
- r-base
- r-base-core
- libxml2-dev
- procps
- libssl-dev
- type: r
cran: [ optparse, ggplot2, RColorBrewer, pheatmap, stringr, matrixStats ]
bioc: [ DESeq2 ]

View File

@@ -8,7 +8,8 @@ par <- list(
deseq2_output = "deseq2",
pca_multiqc = "pca.vals_mqc.tsv",
dists_multiqc = "sample.dists_mqc.tsv",
vst = FALSE
vst = FALSE,
outdir = '.'
)
meta <- list(
resources_dir = "src/deseq2_qc"
@@ -100,7 +101,7 @@ saveRDS(dds, file = sub("\\.dds\\.RData$", ".rds", DDSFile))
##' @author Gavin Kelly
plotPCA_vst <- function(object, ntop = 500, assay = length(assays(object))) {
rv <- rowVars(assay(object, assay))
rv <- rowVars(assay(object, assay), useNames = TRUE)
select <- order(rv, decreasing = TRUE)[seq_len(min(ntop, length(rv)))]
pca <- prcomp(t(assay(object, assay)[select, ]), center = TRUE, scale = FALSE)
percentVar <- pca$sdev^2 / sum(pca$sdev^2)

View File

@@ -0,0 +1,32 @@
name: "copy_if_exists"
argument_groups:
- name: "Input"
arguments:
- name: "--required_file"
type: file
must_exist: false
required: true
example: /tmp/rnaseq_workflow_config/required_file.txt
- name: --optional_file
type: file
must_exist: false
example: /tmp/rnaseq_workflow_config/optional_file.txt
- name: "Ouput"
arguments:
- name: "--output"
type: file
direction: output
default: copy_if_exists_output
resources:
- type: bash_script
path: script.sh
- path: /src/assets/required_file.txt
- path: /src/assets/optional_file.txt
engines:
- type: docker
image: ubuntu:22.04
runners:
- type: executable
- type: nextflow

View File

@@ -0,0 +1,25 @@
#!/bin/bash
set -eo pipefail
mkdir -p $par_output
# This file is checked by the Nextflow module wrapper
cp $par_required_file "$par_output"
# If the variable is empty, we use the default one (registered as a resource)
if [ -z $par_optional_file ]; then
echo "No optional_file provided, using the default"
cp $meta_resources_dir/optional_file.txt "$par_output"
else
echo "Optional file provided"
if [ -f $par_optional_file ]; then
cp $par_optional_file "$par_output"
else
# Unreachable: the Viash-generated module checks this
echo "Optional file does not exist"
exit 1
fi
fi
echo "Done"

View File

@@ -39,10 +39,8 @@ resources:
engines:
- type: docker
image: ubuntu:22.04
image: rocker/r2u:22.04
setup:
- type: apt
packages: [ r-base, libcurl4-openssl-dev ]
- type: r
bioc: [ SummarizedExperiment, tximeta ]
runners:

View File

@@ -267,6 +267,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -367,16 +370,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/bbmap/bbmap_bbsplit"
executable: "target/nextflow/bbmap/bbmap_bbsplit/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// bbmap_bbsplit main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3101,6 +3327,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3219,16 +3449,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/bbmap/bbmap_bbsplit",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3729,7 +3959,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3743,6 +3973,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -187,10 +187,10 @@
"index": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.index.index`, example: `BBSplit_index`. Location to write the index",
"help_text": "Type: `file`, default: `$id.$key.index.index`, example: `BBSplit_index`. Location to write the index.\n"
"description": "Type: `file`, default: `$id.$key.index`, example: `BBSplit_index`. Location to write the index",
"help_text": "Type: `file`, default: `$id.$key.index`, example: `BBSplit_index`. Location to write the index.\n"
,
"default":"$id.$key.index.index"
"default":"$id.$key.index"
}

View File

@@ -226,6 +226,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -336,16 +339,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/bedtools/bedtools_genomecov"
executable: "target/nextflow/bedtools/bedtools_genomecov/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// bedtools_genomecov main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3073,6 +3299,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3206,16 +3436,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/bedtools/bedtools_genomecov",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3677,7 +3907,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3691,6 +3921,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -982,6 +982,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -1082,16 +1085,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/fastp"
executable: "target/nextflow/fastp/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// fastp main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3901,6 +4127,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -4022,16 +4252,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/fastp",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4606,7 +4836,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4620,6 +4850,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -102,10 +102,10 @@
"overlapped_out": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.overlapped_out.overlapped_out`. For each read pair, output the overlapped region if it has no any mismatched base",
"help_text": "Type: `file`, default: `$id.$key.overlapped_out.overlapped_out`. For each read pair, output the overlapped region if it has no any mismatched base.\n"
"description": "Type: `file`, default: `$id.$key.overlapped_out`. For each read pair, output the overlapped region if it has no any mismatched base",
"help_text": "Type: `file`, default: `$id.$key.overlapped_out`. For each read pair, output the overlapped region if it has no any mismatched base.\n"
,
"default":"$id.$key.overlapped_out.overlapped_out"
"default":"$id.$key.overlapped_out"
}

View File

@@ -239,6 +239,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -339,16 +342,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/fastqc"
executable: "target/nextflow/fastqc/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// fastqc main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3062,6 +3288,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3181,16 +3411,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/fastqc",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3678,7 +3908,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3692,6 +3922,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -543,6 +543,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -644,16 +647,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/featurecounts"
executable: "target/nextflow/featurecounts/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// featurecounts main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3427,6 +3653,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3548,16 +3778,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/featurecounts",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3662,7 +3892,7 @@ $( if [ ! -z ${VIASH_META_MEMORY_PIB+x} ]; then echo "${VIASH_META_MEMORY_PIB}"
## VIASH END
# create temporary directory
tmp_dir=\\$(mktemp -d -p "\\$meta_temp_dir" "\\${meta_functionality_name}_XXXXXX")
tmp_dir=\\$(mktemp -d -p "\\$meta_temp_dir" "\\${meta_name}_XXXXXX")
mkdir -p "\\$tmp_dir/temp"
# create detailed_results directory if variable is set and directory does not exist
@@ -4089,7 +4319,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4103,6 +4333,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -573,10 +573,10 @@
"detailed_results": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.detailed_results.detailed_results`, example: `detailed_results`. Directory to save the detailed assignment results",
"help_text": "Type: `file`, default: `$id.$key.detailed_results.detailed_results`, example: `detailed_results`. Directory to save the detailed assignment results. Use `--detailed_results_format` to determine the format of the detailed results.\n"
"description": "Type: `file`, default: `$id.$key.detailed_results`, example: `detailed_results`. Directory to save the detailed assignment results",
"help_text": "Type: `file`, default: `$id.$key.detailed_results`, example: `detailed_results`. Directory to save the detailed assignment results. Use `--detailed_results_format` to determine the format of the detailed results.\n"
,
"default":"$id.$key.detailed_results.detailed_results"
"default":"$id.$key.detailed_results"
}

View File

@@ -88,6 +88,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -189,16 +192,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/fq_subsample"
executable: "target/nextflow/fq_subsample/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// fq_subsample main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2914,6 +3140,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3031,16 +3261,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/fq_subsample",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3459,7 +3689,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3473,6 +3703,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -47,10 +47,10 @@
"output_1": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_1.output_1`. Sampled read 1 fastq files",
"help_text": "Type: `file`, default: `$id.$key.output_1.output_1`. Sampled read 1 fastq files. Output will be gzipped if ends in `.gz`."
"description": "Type: `file`, default: `$id.$key.output_1`. Sampled read 1 fastq files",
"help_text": "Type: `file`, default: `$id.$key.output_1`. Sampled read 1 fastq files. Output will be gzipped if ends in `.gz`."
,
"default":"$id.$key.output_1.output_1"
"default":"$id.$key.output_1"
}
@@ -58,10 +58,10 @@
"output_2": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_2.output_2`. Sampled read 2 fastq files",
"help_text": "Type: `file`, default: `$id.$key.output_2.output_2`. Sampled read 2 fastq files. Output will be gzipped if ends in `.gz`."
"description": "Type: `file`, default: `$id.$key.output_2`. Sampled read 2 fastq files",
"help_text": "Type: `file`, default: `$id.$key.output_2`. Sampled read 2 fastq files. Output will be gzipped if ends in `.gz`."
,
"default":"$id.$key.output_2.output_2"
"default":"$id.$key.output_2"
}

View File

@@ -582,6 +582,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -684,16 +687,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/gffread"
executable: "target/nextflow/gffread/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// gffread main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3482,6 +3708,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3605,16 +3835,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/gffread",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4192,7 +4422,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4206,6 +4436,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -113,6 +113,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -217,16 +220,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/kallisto/kallisto_index"
executable: "target/nextflow/kallisto/kallisto_index/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// kallisto_index main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2948,6 +3174,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3070,16 +3300,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/kallisto/kallisto_index",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3509,7 +3739,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3523,6 +3753,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -47,10 +47,10 @@
"index": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.index.index`, example: `Kallisto_index`. ",
"help_text": "Type: `file`, default: `$id.$key.index.index`, example: `Kallisto_index`. "
"description": "Type: `file`, default: `$id.$key.index`, example: `Kallisto_index`. ",
"help_text": "Type: `file`, default: `$id.$key.index`, example: `Kallisto_index`. "
,
"default":"$id.$key.index.index"
"default":"$id.$key.index"
}

View File

@@ -137,6 +137,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -245,16 +248,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/kallisto/kallisto_quant"
executable: "target/nextflow/kallisto/kallisto_quant/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// kallisto_quant main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2975,6 +3201,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3104,16 +3334,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/kallisto/kallisto_quant",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3557,7 +3787,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3571,6 +3801,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -47,10 +47,10 @@
"output_dir": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.output_dir.output_dir`. Directory to write output to",
"help_text": "Type: `file`, required, default: `$id.$key.output_dir.output_dir`. Directory to write output to."
"description": "Type: `file`, required, default: `$id.$key.output_dir`. Directory to write output to",
"help_text": "Type: `file`, required, default: `$id.$key.output_dir`. Directory to write output to."
,
"default":"$id.$key.output_dir.output_dir"
"default":"$id.$key.output_dir"
}
@@ -58,10 +58,10 @@
"log": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.log.log`. File containing log information from running kallisto quant",
"help_text": "Type: `file`, default: `$id.$key.log.log`. File containing log information from running kallisto quant"
"description": "Type: `file`, default: `$id.$key.log`. File containing log information from running kallisto quant",
"help_text": "Type: `file`, default: `$id.$key.log`. File containing log information from running kallisto quant"
,
"default":"$id.$key.log.log"
"default":"$id.$key.log"
}

View File

@@ -357,6 +357,9 @@ info:
doi: "10.1093/bioinformatics/btw354"
licence: "GPL v3 or later"
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -455,16 +458,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/multiqc"
executable: "target/nextflow/multiqc/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// multiqc main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3246,6 +3472,10 @@ meta = [
"licence" : "GPL v3 or later"
},
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3365,16 +3595,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/multiqc",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3935,7 +4165,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3949,6 +4179,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -48,10 +48,10 @@
"output_data": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_data.output_data`, example: `multiqc_data`. Output directory for parsed data files",
"help_text": "Type: `file`, default: `$id.$key.output_data.output_data`, example: `multiqc_data`. Output directory for parsed data files. If not provided, parsed data will not be published.\n"
"description": "Type: `file`, default: `$id.$key.output_data`, example: `multiqc_data`. Output directory for parsed data files",
"help_text": "Type: `file`, default: `$id.$key.output_data`, example: `multiqc_data`. Output directory for parsed data files. If not provided, parsed data will not be published.\n"
,
"default":"$id.$key.output_data.output_data"
"default":"$id.$key.output_data"
}
@@ -59,10 +59,10 @@
"output_plots": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_plots.output_plots`, example: `multiqc_plots`. Output directory for generated plots",
"help_text": "Type: `file`, default: `$id.$key.output_plots.output_plots`, example: `multiqc_plots`. Output directory for generated plots. If not provided, plots will not be published.\n"
"description": "Type: `file`, default: `$id.$key.output_plots`, example: `multiqc_plots`. Output directory for generated plots",
"help_text": "Type: `file`, default: `$id.$key.output_plots`, example: `multiqc_plots`. Output directory for generated plots. If not provided, plots will not be published.\n"
,
"default":"$id.$key.output_plots.output_plots"
"default":"$id.$key.output_plots"
}

View File

@@ -161,6 +161,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -263,16 +266,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/qualimap/qualimap_rnaseq"
executable: "target/nextflow/qualimap/qualimap_rnaseq/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// qualimap_rnaseq main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3005,6 +3231,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3128,16 +3358,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/qualimap/qualimap_rnaseq",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3588,7 +3818,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3602,6 +3832,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -58,10 +58,10 @@
"counts": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.counts.counts`. Output file for computed counts",
"help_text": "Type: `file`, default: `$id.$key.counts.counts`. Output file for computed counts."
"description": "Type: `file`, default: `$id.$key.counts`. Output file for computed counts",
"help_text": "Type: `file`, default: `$id.$key.counts`. Output file for computed counts."
,
"default":"$id.$key.counts.counts"
"default":"$id.$key.counts"
}

View File

@@ -724,6 +724,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -851,16 +854,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/rsem/rsem_calculate_expression"
executable: "target/nextflow/rsem/rsem_calculate_expression/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// rsem_calculate_expression main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3514,6 +3740,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3660,16 +3890,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/rsem/rsem_calculate_expression",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4221,7 +4451,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4235,6 +4465,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -268,6 +268,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -415,16 +418,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/rsem/rsem_prepare_reference"
executable: "target/nextflow/rsem/rsem_prepare_reference/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// rsem_prepare_reference main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3085,6 +3311,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3244,16 +3474,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/rsem/rsem_prepare_reference",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3705,7 +3935,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3719,6 +3949,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -47,10 +47,10 @@
"output": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.output.output`. Directory containing reference files generated by RSEM",
"help_text": "Type: `file`, required, default: `$id.$key.output.output`. Directory containing reference files generated by RSEM."
"description": "Type: `file`, required, default: `$id.$key.output`. Directory containing reference files generated by RSEM",
"help_text": "Type: `file`, required, default: `$id.$key.output`. Directory containing reference files generated by RSEM."
,
"default":"$id.$key.output.output"
"default":"$id.$key.output"
}

View File

@@ -68,6 +68,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -174,16 +177,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/rseqc/rseqc_bamstat"
executable: "target/nextflow/rseqc/rseqc_bamstat/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// rseqc_bamstat main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2905,6 +3131,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3035,16 +3265,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/rseqc/rseqc_bamstat",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3445,7 +3675,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3459,6 +3689,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -47,10 +47,10 @@
"output": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output.output`. Output file (txt) with mapping quality statistics",
"help_text": "Type: `file`, default: `$id.$key.output.output`. Output file (txt) with mapping quality statistics."
"description": "Type: `file`, default: `$id.$key.output`. Output file (txt) with mapping quality statistics",
"help_text": "Type: `file`, default: `$id.$key.output`. Output file (txt) with mapping quality statistics."
,
"default":"$id.$key.output.output"
"default":"$id.$key.output"
}

View File

@@ -96,6 +96,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -200,16 +203,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/rseqc/rseqc_inferexperiment"
executable: "target/nextflow/rseqc/rseqc_inferexperiment/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// rseqc_inferexperiment main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2942,6 +3168,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3068,16 +3298,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/rseqc/rseqc_inferexperiment",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3481,7 +3711,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3495,6 +3725,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -185,6 +185,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -293,16 +296,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/rseqc/rseqc_inner_distance"
executable: "target/nextflow/rseqc/rseqc_inner_distance/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// rseqc_inner_distance main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3035,6 +3261,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3168,16 +3398,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/rseqc/rseqc_inner_distance",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3604,7 +3834,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3618,6 +3848,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -107,10 +107,10 @@
"output_stats": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_stats.output_stats`. output file (txt) with summary statistics of inner distances of paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_stats.output_stats`. output file (txt) with summary statistics of inner distances of paired reads"
"description": "Type: `file`, default: `$id.$key.output_stats`. output file (txt) with summary statistics of inner distances of paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_stats`. output file (txt) with summary statistics of inner distances of paired reads"
,
"default":"$id.$key.output_stats.output_stats"
"default":"$id.$key.output_stats"
}
@@ -118,10 +118,10 @@
"output_dist": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_dist.output_dist`. output file (txt) with inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_dist.output_dist`. output file (txt) with inner distances of all paired reads"
"description": "Type: `file`, default: `$id.$key.output_dist`. output file (txt) with inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_dist`. output file (txt) with inner distances of all paired reads"
,
"default":"$id.$key.output_dist.output_dist"
"default":"$id.$key.output_dist"
}
@@ -129,10 +129,10 @@
"output_freq": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_freq.output_freq`. output file (txt) with frequencies of inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_freq.output_freq`. output file (txt) with frequencies of inner distances of all paired reads"
"description": "Type: `file`, default: `$id.$key.output_freq`. output file (txt) with frequencies of inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_freq`. output file (txt) with frequencies of inner distances of all paired reads"
,
"default":"$id.$key.output_freq.output_freq"
"default":"$id.$key.output_freq"
}
@@ -140,10 +140,10 @@
"output_plot": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_plot.output_plot`. output file (pdf) with histogram plot of of inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_plot.output_plot`. output file (pdf) with histogram plot of of inner distances of all paired reads"
"description": "Type: `file`, default: `$id.$key.output_plot`. output file (pdf) with histogram plot of of inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_plot`. output file (pdf) with histogram plot of of inner distances of all paired reads"
,
"default":"$id.$key.output_plot.output_plot"
"default":"$id.$key.output_plot"
}
@@ -151,10 +151,10 @@
"output_plot_r": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output_plot_r.output_plot_r`. output file (R) with script of histogram plot of of inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_plot_r.output_plot_r`. output file (R) with script of histogram plot of of inner distances of all paired reads"
"description": "Type: `file`, default: `$id.$key.output_plot_r`. output file (R) with script of histogram plot of of inner distances of all paired reads",
"help_text": "Type: `file`, default: `$id.$key.output_plot_r`. output file (R) with script of histogram plot of of inner distances of all paired reads"
,
"default":"$id.$key.output_plot_r.output_plot_r"
"default":"$id.$key.output_plot_r"
}

View File

@@ -176,6 +176,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -276,16 +279,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/salmon/salmon_index"
executable: "target/nextflow/salmon/salmon_index/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// salmon_index main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3007,6 +3233,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3128,16 +3358,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/salmon/salmon_index",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3224,7 +3454,7 @@ for par in \\${unset_if_false[@]}; do
[[ "\\$test_val" == "false" ]] && unset \\$par
done
tmp_dir=\\$(mktemp -d -p "\\$meta_temp_dir" "\\${meta_functionality_name}_XXXXXX")
tmp_dir=\\$(mktemp -d -p "\\$meta_temp_dir" "\\${meta_name}_XXXXXX")
mkdir -p "\\$tmp_dir/temp"
if [[ -f "\\$par_genome" ]] && [[ ! "\\$par_decoys" ]]; then
@@ -3593,7 +3823,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3607,6 +3837,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -153,10 +153,10 @@
"index": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.index.index`, example: `Salmon_index`. Salmon index\n",
"help_text": "Type: `file`, required, default: `$id.$key.index.index`, example: `Salmon_index`. Salmon index\n"
"description": "Type: `file`, required, default: `$id.$key.index`, example: `Salmon_index`. Salmon index\n",
"help_text": "Type: `file`, required, default: `$id.$key.index`, example: `Salmon_index`. Salmon index\n"
,
"default":"$id.$key.index.index"
"default":"$id.$key.index"
}

View File

@@ -1072,6 +1072,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -1172,16 +1175,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/salmon/salmon_quant"
executable: "target/nextflow/salmon/salmon_quant/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// salmon_quant main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3842,6 +4068,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3963,16 +4193,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/salmon/salmon_quant",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4608,7 +4838,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4622,6 +4852,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -152,10 +152,10 @@
"output": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.output.output`, example: `quant_output`. Output quantification directory",
"help_text": "Type: `file`, required, default: `$id.$key.output.output`, example: `quant_output`. Output quantification directory.\n"
"description": "Type: `file`, required, default: `$id.$key.output`, example: `quant_output`. Output quantification directory",
"help_text": "Type: `file`, required, default: `$id.$key.output`, example: `quant_output`. Output quantification directory.\n"
,
"default":"$id.$key.output.output"
"default":"$id.$key.output"
}
@@ -584,10 +584,10 @@
"aux_dir": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.aux_dir.aux_dir`, example: `aux_info`. The sub-directory of the quantification directory where auxiliary information e",
"help_text": "Type: `file`, default: `$id.$key.aux_dir.aux_dir`, example: `aux_info`. The sub-directory of the quantification directory where auxiliary information e.g. bootstraps, bias parameters, etc. will be written.\n"
"description": "Type: `file`, default: `$id.$key.aux_dir`, example: `aux_info`. The sub-directory of the quantification directory where auxiliary information e",
"help_text": "Type: `file`, default: `$id.$key.aux_dir`, example: `aux_info`. The sub-directory of the quantification directory where auxiliary information e.g. bootstraps, bias parameters, etc. will be written.\n"
,
"default":"$id.$key.aux_dir.aux_dir"
"default":"$id.$key.aux_dir"
}

View File

@@ -66,6 +66,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -172,16 +175,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/samtools/samtools_flagstat"
executable: "target/nextflow/samtools/samtools_flagstat/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// samtools_flagstat main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2901,6 +3127,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3027,16 +3257,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/samtools/samtools_flagstat",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3437,7 +3667,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3451,6 +3681,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -75,6 +75,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -182,16 +185,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/samtools/samtools_idxstats"
executable: "target/nextflow/samtools/samtools_idxstats/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// samtools_idxstats main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2912,6 +3138,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3039,16 +3269,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/samtools/samtools_idxstats",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3447,7 +3677,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3461,6 +3691,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -84,6 +84,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -188,16 +191,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/samtools/samtools_index"
executable: "target/nextflow/samtools/samtools_index/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// samtools_index main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2928,6 +3154,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3052,16 +3282,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/samtools/samtools_index",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3471,7 +3701,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3485,6 +3715,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -227,6 +227,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -331,16 +334,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/samtools/samtools_sort"
executable: "target/nextflow/samtools/samtools_sort/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// samtools_sort main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3100,6 +3326,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3224,16 +3454,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/samtools/samtools_sort",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3691,7 +3921,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3705,6 +3935,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -295,6 +295,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -400,16 +403,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/samtools/samtools_stats"
executable: "target/nextflow/samtools/samtools_stats/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// samtools_stats main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3169,6 +3395,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3294,16 +3524,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/samtools/samtools_stats",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3753,7 +3983,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3767,6 +3997,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -487,6 +487,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -593,16 +596,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/sortmerna"
executable: "target/nextflow/sortmerna/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// sortmerna main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3333,6 +3559,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3459,16 +3689,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/sortmerna",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4003,7 +4233,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4017,6 +4247,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -79,10 +79,10 @@
"output": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.output.output`. Directory and file prefix for aligned output",
"help_text": "Type: `file`, default: `$id.$key.output.output`. Directory and file prefix for aligned output. The appropriate extension: \n(fasta|fastq|blast|sam|etc) is automatically added.\nIf \u0027dir\u0027 is not specified, the output is created in the WORKDIR/out/.\nIf \u0027pfx\u0027 is not specified, the prefix \u0027aligned\u0027 is used.\n"
"description": "Type: `file`, default: `$id.$key.output`. Directory and file prefix for aligned output",
"help_text": "Type: `file`, default: `$id.$key.output`. Directory and file prefix for aligned output. The appropriate extension: \n(fasta|fastq|blast|sam|etc) is automatically added.\nIf \u0027dir\u0027 is not specified, the output is created in the WORKDIR/out/.\nIf \u0027pfx\u0027 is not specified, the prefix \u0027aligned\u0027 is used.\n"
,
"default":"$id.$key.output.output"
"default":"$id.$key.output"
}
@@ -90,10 +90,10 @@
"other": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.other.other`. Create Non-aligned reads output file with this path/prefix",
"help_text": "Type: `file`, default: `$id.$key.other.other`. Create Non-aligned reads output file with this path/prefix. Must be used with fastx."
"description": "Type: `file`, default: `$id.$key.other`. Create Non-aligned reads output file with this path/prefix",
"help_text": "Type: `file`, default: `$id.$key.other`. Create Non-aligned reads output file with this path/prefix. Must be used with fastx."
,
"default":"$id.$key.other.other"
"default":"$id.$key.other"
}

View File

@@ -2540,6 +2540,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -2662,16 +2665,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/star/star_align_reads"
executable: "target/nextflow/star/star_align_reads/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// star_align_reads main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -86,64 +86,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -155,10 +147,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -177,7 +172,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -196,15 +191,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -217,6 +205,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1670,6 +1668,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1727,8 +1881,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1741,7 +1893,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1753,33 +1905,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1810,13 +1946,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1833,7 +1966,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1864,13 +1997,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1878,18 +2007,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2563,7 +2691,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2720,12 +2849,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2738,19 +2891,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2759,23 +2987,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -5794,6 +6020,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -5942,16 +6172,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/star/star_align_reads",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -5982,7 +6212,7 @@ meta = [
// inner workflow hook
def innerWorkflowFactory(args) {
def rawScript = '''set -e
tempscript=".viash_script.sh"
tempscript=".viash_script.py"
cat > "$tempscript" << VIASHMAIN
import tempfile
import subprocess
@@ -6645,7 +6875,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -6659,6 +6889,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -221,6 +221,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -332,16 +335,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/star/star_genome_generate"
executable: "target/nextflow/star/star_genome_generate/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// star_genome_generate main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3063,6 +3289,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3194,16 +3424,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/star/star_genome_generate",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3292,10 +3522,10 @@ STAR \\\\
\\${par_genome_sa_index_nbases:+--genomeSAindexNbases "\\${par_genome_sa_index_nbases}"} \\\\
\\${par_sjdb_gtf_chr_prefix:+--sjdbGTFchrPrefix "\\${par_sjdb_gtf_chr_prefix}"} \\\\
\\${par_sjdb_gtf_feature_exon:+--sjdbGTFfeatureExon "\\${par_sjdb_gtf_feature_exon}"} \\\\
\\${par_sjdb_gtf_tag_exon_parent_transcript:+--sjdbGTFtag_exon_parent_transcript "\\${par_sjdb_gtf_tag_exon_parent_transcript}"} \\\\
\\${par_sjdb_gtf_tag_exon_parent_gene:+--sjdbGTFtag_exon_parent_gene "\\${par_sjdb_gtf_tag_exon_parent_gene}"} \\\\
\\${par_sjdb_gtf_tag_exon_parent_geneName:+--sjdbGTFtag_exon_parent_geneName "\\${par_sjdb_gtf_tag_exon_parent_geneName}"} \\\\
\\${par_sjdb_gtf_tag_exon_parent_geneType:+--sjdbGTFtag_exon_parent_geneType "\\${sjdbGTFtag_exon_parent_geneType}"} \\\\
\\${par_sjdb_gtf_tag_exon_parent_transcript:+--sjdbGTFtagExonParentTranscript "\\${par_sjdb_gtf_tag_exon_parent_transcript}"} \\\\
\\${par_sjdb_gtf_tag_exon_parent_gene:+--sjdbGTFtagExonParentGene "\\${par_sjdb_gtf_tag_exon_parent_gene}"} \\\\
\\${sjdb_gtf_tag_exon_parent_gene_name:+--sjdbGTFtagExonParentGeneName "\\${sjdb_gtf_tag_exon_parent_gene_name}"} \\\\
\\${sjdb_gtf_tag_exon_parent_gene_type:+--sjdbGTFtagExonParentGeneType "\\${sjdb_gtf_tag_exon_parent_gene_type}"} \\\\
\\${par_limit_genome_generate_ram:+--limitGenomeGenerateRAM "\\${par_limit_genome_generate_ram}"} \\\\
\\${par_genome_chr_bin_nbits:+--genomeChrBinNbits "\\${par_genome_chr_bin_nbits}"} \\\\
\\${par_genome_sa_sparse_d:+--genomeSAsparseD "\\${par_genome_sa_sparse_d}"} \\\\
@@ -3636,7 +3866,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3650,6 +3880,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -187,10 +187,10 @@
"index": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.index.index`. STAR index directory",
"help_text": "Type: `file`, required, default: `$id.$key.index.index`. STAR index directory."
"description": "Type: `file`, required, default: `STAR_index`. STAR index directory",
"help_text": "Type: `file`, required, default: `STAR_index`. STAR index directory."
,
"default":"$id.$key.index.index"
"default":"STAR_index"
}

View File

@@ -668,6 +668,9 @@ test_resources:
is_executable: true
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -769,16 +772,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/trimgalore"
executable: "target/nextflow/trimgalore/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// trimgalore main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3438,6 +3664,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3559,16 +3789,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/trimgalore",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4138,7 +4368,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4152,6 +4382,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -527,10 +527,10 @@
"output_dir": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.output_dir.output_dir`. If specified all output will be written to this directory instead of the current directory",
"help_text": "Type: `file`, required, default: `$id.$key.output_dir.output_dir`. If specified all output will be written to this directory instead of the current directory."
"description": "Type: `file`, required, default: `trimmed_output`. If specified all output will be written to this directory instead of the current directory",
"help_text": "Type: `file`, required, default: `trimmed_output`. If specified all output will be written to this directory instead of the current directory."
,
"default":"$id.$key.output_dir.output_dir"
"default":"trimmed_output"
}

View File

@@ -509,6 +509,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -610,16 +613,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/umi_tools/umi_tools_dedup"
executable: "target/nextflow/umi_tools/umi_tools_dedup/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// umi_tools_dedup main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -85,64 +85,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -154,10 +146,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -176,7 +171,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -195,15 +190,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -216,6 +204,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1669,6 +1667,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1726,8 +1880,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1740,7 +1892,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1752,33 +1904,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1809,13 +1945,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1832,7 +1965,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1863,13 +1996,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1877,18 +2006,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2562,7 +2690,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2719,12 +2848,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2737,19 +2890,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2758,23 +2986,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3364,6 +3590,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3486,16 +3716,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/umi_tools/umi_tools_dedup",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -4008,7 +4238,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -4022,6 +4252,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -68,10 +68,10 @@
"output": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.output.output`. Deduplicated BAM file",
"help_text": "Type: `file`, required, default: `$id.$key.output.output`. Deduplicated BAM file."
"description": "Type: `file`, required, default: `$id.$key.output`. Deduplicated BAM file",
"help_text": "Type: `file`, required, default: `$id.$key.output`. Deduplicated BAM file."
,
"default":"$id.$key.output.output"
"default":"$id.$key.output"
}

View File

@@ -331,6 +331,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -433,16 +436,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/umi_tools/umi_tools_extract"
executable: "target/nextflow/umi_tools/umi_tools_extract/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// umi_tools_extract main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -3175,6 +3401,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3298,16 +3528,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/umi_tools/umi_tools_extract",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3809,7 +4039,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3823,6 +4053,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -67,10 +67,10 @@
"output": {
"type":
"string",
"description": "Type: `file`, required, default: `$id.$key.output.output`. Output file for read 1",
"help_text": "Type: `file`, required, default: `$id.$key.output.output`. Output file for read 1."
"description": "Type: `file`, required, default: `$id.$key.output`. Output file for read 1",
"help_text": "Type: `file`, required, default: `$id.$key.output`. Output file for read 1."
,
"default":"$id.$key.output.output"
"default":"$id.$key.output"
}
@@ -78,10 +78,10 @@
"read2_out": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.read2_out.read2_out`. Output file for read 2",
"help_text": "Type: `file`, default: `$id.$key.read2_out.read2_out`. Output file for read 2."
"description": "Type: `file`, default: `$id.$key.read2_out`. Output file for read 2",
"help_text": "Type: `file`, default: `$id.$key.read2_out`. Output file for read 2."
,
"default":"$id.$key.read2_out.read2_out"
"default":"$id.$key.read2_out"
}
@@ -269,10 +269,10 @@
"log": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.log.log`. File with logging information",
"help_text": "Type: `file`, default: `$id.$key.log.log`. File with logging information."
"description": "Type: `file`, default: `$id.$key.log`. File with logging information",
"help_text": "Type: `file`, default: `$id.$key.log`. File with logging information."
,
"default":"$id.$key.log.log"
"default":"$id.$key.log"
}
@@ -301,10 +301,10 @@
"error": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.error.error`. File with error information",
"help_text": "Type: `file`, default: `$id.$key.error.error`. File with error information."
"description": "Type: `file`, default: `$id.$key.error`. File with error information",
"help_text": "Type: `file`, default: `$id.$key.error`. File with error information."
,
"default":"$id.$key.error.error"
"default":"$id.$key.error"
}
@@ -332,10 +332,10 @@
"timeit": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.timeit.timeit`. Store timing information in file",
"help_text": "Type: `file`, default: `$id.$key.timeit.timeit`. Store timing information in file."
"description": "Type: `file`, default: `$id.$key.timeit`. Store timing information in file",
"help_text": "Type: `file`, default: `$id.$key.timeit`. Store timing information in file."
,
"default":"$id.$key.timeit.timeit"
"default":"$id.$key.timeit"
}

View File

@@ -153,6 +153,9 @@ test_resources:
path: "test_data"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -255,16 +258,16 @@ build_info:
engine: "docker|native"
output: "target/nextflow/umi_tools/umi_tools_prepareforrsem"
executable: "target/nextflow/umi_tools/umi_tools_prepareforrsem/main.nf"
viash_version: "0.9.0"
git_commit: "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b"
git_remote: "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox"
git_tag: "v0.2.0-27-g952ff08"
viash_version: "0.9.4"
git_commit: "c733de2e4cc21eccb241060f82bce0b332f79d81"
git_remote: "https://github.com/viash-hub/biobox"
git_tag: "v0.2.0-32-gc733de2"
package_config:
name: "biobox"
version: "main"
description: "A collection of bioinformatics tools for working with sequence data.\n"
info: null
viash_version: "0.9.0"
viash_version: "0.9.4"
source: "src"
target: "target"
config_mods:

View File

@@ -1,6 +1,6 @@
// umi_tools_prepareforrsem main
//
// This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
// This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
// work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
// Intuitive.
//
@@ -82,64 +82,56 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
foundClass = "List[${e.foundClass}]"
}
} else if (par.type == "string") {
// cast to string if need be
// cast to string if need be. only cast if the value is a GString
if (value instanceof GString) {
value = value.toString()
value = value as String
}
expectedClass = value instanceof String ? null : "String"
} else if (par.type == "integer") {
// cast to integer if need be
if (value instanceof String) {
if (value !instanceof Integer) {
try {
value = value.toInteger()
value = value as Integer
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Integer"
}
}
if (value instanceof java.math.BigInteger) {
value = value.intValue()
}
expectedClass = value instanceof Integer ? null : "Integer"
} else if (par.type == "long") {
// cast to long if need be
if (value instanceof String) {
if (value !instanceof Long) {
try {
value = value.toLong()
value = value as Long
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Long"
}
}
if (value instanceof Integer) {
value = value.toLong()
}
expectedClass = value instanceof Long ? null : "Long"
} else if (par.type == "double") {
// cast to double if need be
if (value instanceof String) {
if (value !instanceof Double) {
try {
value = value.toDouble()
value = value as Double
} catch (NumberFormatException e) {
// do nothing
expectedClass = "Double"
}
}
if (value instanceof java.math.BigDecimal) {
value = value.doubleValue()
} else if (par.type == "float") {
// cast to float if need be
if (value !instanceof Float) {
try {
value = value as Float
} catch (NumberFormatException e) {
expectedClass = "Float"
}
}
if (value instanceof Float) {
value = value.toDouble()
}
expectedClass = value instanceof Double ? null : "Double"
} else if (par.type == "boolean" | par.type == "boolean_true" | par.type == "boolean_false") {
// cast to boolean if need be
if (value instanceof String) {
def valueLower = value.toLowerCase()
if (valueLower == "true") {
value = true
} else if (valueLower == "false") {
value = false
if (value !instanceof Boolean) {
try {
value = value as Boolean
} catch (Exception e) {
expectedClass = "Boolean"
}
}
expectedClass = value instanceof Boolean ? null : "Boolean"
} else if (par.type == "file" && (par.direction == "input" || stage == "output")) {
// cast to path if need be
if (value instanceof String) {
@@ -151,10 +143,13 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
expectedClass = value instanceof Path ? null : "Path"
} else if (par.type == "file" && stage == "input" && par.direction == "output") {
// cast to string if need be
if (value instanceof GString) {
value = value.toString()
if (value !instanceof String) {
try {
value = value as String
} catch (Exception e) {
expectedClass = "String"
}
}
expectedClass = value instanceof String ? null : "String"
} else {
// didn't find a match for par.type
expectedClass = par.type
@@ -173,7 +168,7 @@ def _checkArgumentType(String stage, Map par, Object value, String errorIdentifi
Map _processInputValues(Map inputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.required) {
if (arg.required && arg.direction == "input") {
assert inputs.containsKey(arg.plainName) && inputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required input argument '${arg.plainName}' is missing"
}
@@ -192,15 +187,8 @@ Map _processInputValues(Map inputs, Map config, String id, String key) {
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/arguments/_processOutputValues.nf'
Map _processOutputValues(Map outputs, Map config, String id, String key) {
Map _checkValidOutputArgument(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
outputs = outputs.collectEntries { name, value ->
def par = config.allArguments.find { it.plainName == name && it.direction == "output" }
assert par != null : "Error in module '${key}' id '${id}': '${name}' is not a valid output argument"
@@ -213,6 +201,16 @@ Map _processOutputValues(Map outputs, Map config, String id, String key) {
return outputs
}
void _checkAllRequiredOuputsPresent(Map outputs, Map config, String id, String key) {
if (!workflow.stubRun) {
config.allArguments.each { arg ->
if (arg.direction == "output" && arg.required) {
assert outputs.containsKey(arg.plainName) && outputs.get(arg.plainName) != null :
"Error in module '${key}' id '${id}': required output argument '${arg.plainName}' is missing"
}
}
}
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/channel/IDChecker.nf'
class IDChecker {
final def items = [] as Set
@@ -1666,6 +1664,162 @@ def joinStates(Closure apply_) {
}
return joinStatesWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishFiles.nf'
def publishFiles(Map args) {
def key_ = args.get("key")
assert key_ != null : "publishFiles: key must be specified"
workflow publishFilesWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1]
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
[id_, inputFiles_, outputFilenames_]
}
| publishFilesProc
emit: input_ch
}
return publishFilesWf
}
process publishFilesProc {
// todo: check publishpath?
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
output:
tuple val(id), path{outputFiles}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
}
// this assumes that the state contains no other values other than those specified in the config
def publishFilesByConfig(Map args) {
def config = args.get("config")
assert config != null : "publishFilesByConfig: config must be specified"
def key_ = args.get("key", config.name)
assert key_ != null : "publishFilesByConfig: key must be specified"
workflow publishFilesSimpleWf {
take: input_ch
main:
input_ch
| map { tup ->
def id_ = tup[0]
def state_ = tup[1] // e.g. [output: new File("myoutput.h5ad"), k: 10]
def origState_ = tup[2] // e.g. [output: '$id.$key.foo.h5ad']
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
.collectMany { par ->
def plainName_ = par.plainName
// if the state does not contain the key, it's an
// optional argument for which the component did
// not generate any output OR multiple channels were emitted
// and the output was just not added to using the channel
// that is now being parsed
if (!state_.containsKey(plainName_)) {
return []
}
def value = state_[plainName_]
// if the parameter is not a file, it should be stored
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[inputPath: [], outputFilename: []]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
// that it should not be returned as a state
if (!origState_.containsKey(plainName_)) {
return []
}
def filenameTemplate = origState_[plainName_]
// if the pararameter is multiple: true, fetch the template
if (par.multiple && filenameTemplate instanceof List) {
filenameTemplate = filenameTemplate[0]
}
// instantiate the template
def filename = filenameTemplate
.replaceAll('\\$id', id_)
.replaceAll('\\$\\{id\\}', id_)
.replaceAll('\\$key', key_)
.replaceAll('\\$\\{key\\}', key_)
if (par.multiple) {
// if the parameter is multiple: true, the filename
// should contain a wildcard '*' that is replaced with
// the index of the file
assert filename.contains("*") : "Module '${key_}' id '${id_}': Multiple output files specified, but no wildcard '*' in the filename: ${filename}"
def outputPerFile = value.withIndex().collect{ val, ix ->
def filename_ix = filename.replace("*", ix.toString())
def inputPath = val instanceof File ? val.toPath() : val
[inputPath: inputPath, outputFilename: filename_ix]
}
def transposedOutputs = ["inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
} else {
def value_ = java.nio.file.Paths.get(filename)
def inputPath = value instanceof File ? value.toPath() : value
return [[inputPath: [inputPath], outputFilename: [filename]]]
}
}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
[id_, inputPaths, outputFilenames]
}
| publishFilesProc
emit: input_ch
}
return publishFilesSimpleWf
}
// helper file: 'src/main/resources/io/viash/runners/nextflow/states/publishStates.nf'
def collectFiles(obj) {
if (obj instanceof java.io.File || obj instanceof Path) {
@@ -1723,8 +1877,6 @@ def publishStates(Map args) {
// the input files and the target output filenames
def inputoutputFilenames_ = collectInputOutputPaths(state_, id_ + "." + key_).transpose()
def inputFiles_ = inputoutputFilenames_[0]
def outputFilenames_ = inputoutputFilenames_[1]
def yamlFilename = yamlTemplate_
.replaceAll('\\$id', id_)
@@ -1737,7 +1889,7 @@ def publishStates(Map args) {
// convert state to yaml blob
def yamlBlob_ = toRelativeTaggedYamlBlob([id: id_] + state_, java.nio.file.Paths.get(yamlFilename))
[id_, yamlBlob_, yamlFilename, inputFiles_, outputFilenames_]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -1749,33 +1901,17 @@ process publishStatesProc {
publishDir path: "${getPublishDir()}/", mode: "copy"
tag "$id"
input:
tuple val(id), val(yamlBlob), val(yamlFile), path(inputFiles, stageAs: "_inputfile?/*"), val(outputFiles)
tuple val(id), val(yamlBlob), val(yamlFile)
output:
tuple val(id), path{[yamlFile] + outputFiles}
tuple val(id), path{[yamlFile]}
script:
def copyCommands = [
inputFiles instanceof List ? inputFiles : [inputFiles],
outputFiles instanceof List ? outputFiles : [outputFiles]
]
.transpose()
.collectMany{infile, outfile ->
if (infile.toString() != outfile.toString()) {
[
"[ -d \"\$(dirname '${outfile.toString()}')\" ] || mkdir -p \"\$(dirname '${outfile.toString()}')\"",
"cp -r '${infile.toString()}' '${outfile.toString()}'"
]
} else {
// no need to copy if infile is the same as outfile
[]
}
}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
echo '${yamlBlob}' > '${yamlFile}'
echo "Copying output files to destination folder"
${copyCommands.join("\n ")}
"""
mkdir -p "\$(dirname '${yamlFile}')"
echo "Storing state as yaml"
cat > '${yamlFile}' << HERE
${yamlBlob}
HERE
"""
}
@@ -1806,13 +1942,10 @@ def publishStatesByConfig(Map args) {
.replaceAll('\\$\\{key\\}', key_)
def yamlDir = java.nio.file.Paths.get(yamlFilename).getParent()
// the processed state is a list of [key, value, inputPath, outputFilename] tuples, where
// the processed state is a list of [key, value] tuples, where
// - key is a String
// - value is any object that can be serialized to a Yaml (so a String/Integer/Long/Double/Boolean, a List, a Map, or a Path)
// - inputPath is a List[Path]
// - outputFilename is a List[String]
// - (key, value) are the tuples that will be saved to the state.yaml file
// - (inputPath, outputFilename) are the files that will be copied from src to dest (relative to the state.yaml)
def processedState =
config.allArguments
.findAll { it.direction == "output" }
@@ -1829,7 +1962,7 @@ def publishStatesByConfig(Map args) {
// in the state as-is, but is not something that needs
// to be copied from the source path to the dest path
if (par.type != "file") {
return [[key: plainName_, value: value, inputPath: [], outputFilename: []]]
return [[key: plainName_, value: value]]
}
// if the orig state does not contain this filename,
// it's an optional argument for which the user specified
@@ -1860,13 +1993,9 @@ def publishStatesByConfig(Map args) {
if (yamlDir != null) {
value_ = yamlDir.relativize(value_)
}
def inputPath = val instanceof File ? val.toPath() : val
[value: value_, inputPath: inputPath, outputFilename: filename_ix]
return value_
}
def transposedOutputs = ["value", "inputPath", "outputFilename"].collectEntries{ key ->
[key, outputPerFile.collect{dic -> dic[key]}]
}
return [[key: plainName_] + transposedOutputs]
return [["key": plainName_, "value": outputPerFile]]
} else {
def value_ = java.nio.file.Paths.get(filename)
// if id contains a slash
@@ -1874,18 +2003,17 @@ def publishStatesByConfig(Map args) {
value_ = yamlDir.relativize(value_)
}
def inputPath = value instanceof File ? value.toPath() : value
return [[key: plainName_, value: value_, inputPath: [inputPath], outputFilename: [filename]]]
return [["key": plainName_, value: value_]]
}
}
def updatedState_ = processedState.collectEntries{[it.key, it.value]}
def inputPaths = processedState.collectMany{it.inputPath}
def outputFilenames = processedState.collectMany{it.outputFilename}
// convert state to yaml blob
def yamlBlob_ = toTaggedYamlBlob([id: id_] + updatedState_)
[id_, yamlBlob_, yamlFilename, inputPaths, outputFilenames]
[id_, yamlBlob_, yamlFilename]
}
| publishStatesProc
emit: input_ch
@@ -2559,7 +2687,8 @@ def _debug(workflowArgs, debugKey) {
def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
def workflowArgs = processWorkflowArgs(args, defaultWfArgs, meta)
def key_ = workflowArgs["key"]
def multipleArgs = meta.config.allArguments.findAll{ it.multiple }.collect{it.plainName}
workflow workflowInstance {
take: input_
@@ -2716,12 +2845,36 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
// TODO: move some of the _meta.join_id wrangling to the safeJoin() function.
def chInitialOutput = chArgsWithDefaults
def chInitialOutputMulti = chArgsWithDefaults
| _debug(workflowArgs, "processed")
// run workflow
| innerWorkflowFactory(workflowArgs)
// check output tuple
| map { id_, output_ ->
def chInitialOutputList = chInitialOutputMulti instanceof List ? chInitialOutputMulti : [chInitialOutputMulti]
assert chInitialOutputList.size() > 0: "should have emitted at least one output channel"
// Add a channel ID to the events, which designates the channel the event was emitted from as a running number
// This number is used to sort the events later when the events are gathered from across the channels.
def chInitialOutputListWithIndexedEvents = chInitialOutputList.withIndex().collect{channel, channelIndex ->
def newChannel = channel
| map {tuple ->
assert tuple instanceof List :
"Error in module '${key_}': element in output channel should be a tuple [id, data, ...otherargs...]\n" +
" Example: [\"id\", [input: file('foo.txt'), arg: 10]].\n" +
" Expected class: List. Found: tuple.getClass() is ${tuple.getClass()}"
def newEvent = [channelIndex] + tuple
return newEvent
}
return newChannel
}
// Put the events into 1 channel, cover case where there is only one channel is emitted
def chInitialOutput = chInitialOutputList.size() > 1 ? \
chInitialOutputListWithIndexedEvents[0].mix(*chInitialOutputListWithIndexedEvents.tail()) : \
chInitialOutputListWithIndexedEvents[0]
def chInitialOutputProcessed = chInitialOutput
| map { tuple ->
def channelId = tuple[0]
def id_ = tuple[1]
def output_ = tuple[2]
// see if output map contains metadata
def meta_ =
@@ -2734,19 +2887,94 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
output_ = output_.findAll{k, v -> k != "_meta"}
// check value types
output_ = _processOutputValues(output_, meta.config, id_, key_)
output_ = _checkValidOutputArgument(output_, meta.config, id_, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && output_.size() == 1) {
output_ = output_.values()[0]
}
[join_id, id_, output_]
[join_id, channelId, id_, output_]
}
// | view{"chInitialOutput: ${it.take(3)}"}
// join the output [prev_id, channel_id, new_id, output] with the previous state [prev_id, state, ...]
def chPublishWithPreviousState = safeJoin(chInitialOutputProcessed, chRunFiltered, key_)
// input tuple format: [join_id, channel_id, id, output, prev_state, ...]
// output tuple format: [join_id, channel_id, id, new_state, ...]
| map{ tup ->
def new_state = workflowArgs.toState(tup.drop(2).take(3))
tup.take(3) + [new_state] + tup.drop(5)
}
if (workflowArgs.auto.publish == "state") {
def chPublishFiles = chPublishWithPreviousState
// input tuple format: [join_id, channel_id, id, new_state, ...]
// output tuple format: [join_id, channel_id, id, new_state]
| map{ tup ->
tup.take(4)
}
safeJoin(chPublishFiles, chArgsWithDefaults, key_)
// input tuple format: [join_id, channel_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(2).take(3)
}
| publishFilesByConfig(key: key_, config: meta.config)
}
// Join the state from the events that were emitted from different channels
def chJoined = chInitialOutputProcessed
| map {tuple ->
def join_id = tuple[0]
def channel_id = tuple[1]
def id = tuple[2]
def other = tuple.drop(3)
// Below, groupTuple is used to join the events. To make sure resuming a workflow
// keeps working, the output state must be deterministic. This means the state needs to be
// sorted with groupTuple's has a 'sort' argument. This argument can be set to 'hash',
// but hashing the state when it is large can be problematic in terms of performance.
// Therefore, a custom comparator function is provided. We add the channel ID to the
// states so that we can use the channel ID to sort the items.
def stateWithChannelID = [[channel_id] * other.size(), other].transpose()
// A comparator that is provided to groupTuple's 'sort' argument is applied
// to all elements of the event tuple (that is not the 'id'). The comparator
// closure that is used below expects the input to be List. So the join_id and
// channel_id must also be wrapped in a list.
[[join_id], [channel_id], id] + stateWithChannelID
}
| groupTuple(by: 2, sort: {a, b -> a[0] <=> b[0]}, size: chInitialOutputList.size(), remainder: true)
| map {join_ids, _, id, statesWithChannelID ->
// Remove the channel IDs from the states
def states = statesWithChannelID.collect{it[1]}
def newJoinId = join_ids.flatten().unique{a, b -> a <=> b}
assert newJoinId.size() == 1: "Multiple events were emitted for '$id'."
def newJoinIdUnique = newJoinId[0]
// Merge the states from the different channels
def newState = states.inject([:]){ old_state, state_to_add ->
return old_state + state_to_add.collectEntries{k, v ->
if (!multipleArgs.contains(k)) {
// if the key is not a multiple argument, we expect only one value
if (old_state.containsKey(k)) {
assert old_state[k] == v : "ID $id: multiple entries for argument $k were emitted."
}
[k, v]
} else {
// if the key is a multiple argument, append the different values into one list
def prevValue = old_state.getOrDefault(k, [])
def prevValueAsList = prevValue instanceof List ? prevValue : [prevValue]
[k, prevValueAsList + v]
}
}
}
_checkAllRequiredOuputsPresent(newState, meta.config, id, key_)
// simplify output if need be
if (workflowArgs.auto.simplifyOutput && newState.size() == 1) {
newState = newState.values()[0]
}
return [newJoinIdUnique, id, newState]
}
// join the output [prev_id, new_id, output] with the previous state [prev_id, state, ...]
def chNewState = safeJoin(chInitialOutput, chRunFiltered, key_)
def chNewState = safeJoin(chJoined, chRunFiltered, key_)
// input tuple format: [join_id, id, output, prev_state, ...]
// output tuple format: [join_id, id, new_state, ...]
| map{ tup ->
@@ -2755,23 +2983,21 @@ def workflowFactory(Map args, Map defaultWfArgs, Map meta) {
}
if (workflowArgs.auto.publish == "state") {
def chPublish = chNewState
def chPublishStates = chNewState
// input tuple format: [join_id, id, new_state, ...]
// output tuple format: [join_id, id, new_state]
| map{ tup ->
tup.take(3)
}
safeJoin(chPublish, chArgsWithDefaults, key_)
safeJoin(chPublishStates, chArgsWithDefaults, key_)
// input tuple format: [join_id, id, new_state, orig_state, ...]
// output tuple format: [id, new_state, orig_state]
| map { tup ->
tup.drop(1).take(3)
}
}
| publishStatesByConfig(key: key_, config: meta.config)
}
// remove join_id and meta
chReturn = chNewState
| map { tup ->
// input tuple format: [join_id, id, new_state, ...]
@@ -2997,6 +3223,10 @@ meta = [
}
],
"status" : "enabled",
"scope" : {
"image" : "public",
"target" : "public"
},
"requirements" : {
"commands" : [
"ps"
@@ -3120,16 +3350,16 @@ meta = [
"runner" : "nextflow",
"engine" : "docker|native",
"output" : "target/nextflow/umi_tools/umi_tools_prepareforrsem",
"viash_version" : "0.9.0",
"git_commit" : "952ff0843093b538cbfd6fefdecf2e7a0bc9e70b",
"git_remote" : "https://x-access-token:ghs_EwAUAMYJ0K4VBHlAEMs4ZP2OyQYqJM0PSfEO@github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-27-g952ff08"
"viash_version" : "0.9.4",
"git_commit" : "c733de2e4cc21eccb241060f82bce0b332f79d81",
"git_remote" : "https://github.com/viash-hub/biobox",
"git_tag" : "v0.2.0-32-gc733de2"
},
"package_config" : {
"name" : "biobox",
"version" : "main",
"description" : "A collection of bioinformatics tools for working with sequence data.\n",
"viash_version" : "0.9.0",
"viash_version" : "0.9.4",
"source" : "src",
"target" : "target",
"config_mods" : [
@@ -3562,7 +3792,7 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
// create process from temp file
def binding = new nextflow.script.ScriptBinding([:])
def session = nextflow.Nextflow.getSession()
def parser = new nextflow.script.ScriptParser(session)
def parser = _getScriptLoader(session)
.setModule(true)
.setBinding(binding)
def moduleScript = parser.runScript(tempFile)
@@ -3576,6 +3806,27 @@ def _vdsl3ProcessFactory(Map workflowArgs, Map meta, String rawScript) {
return scriptMeta.getProcess(procKey)
}
// use Reflection to get a ScriptParser / ScriptLoader
// <25.02.0-edge: new nextflow.script.ScriptParser(session)
// >=25.02.0-edge: nextflow.script.ScriptLoaderFactory.create(session)
def _getScriptLoader(nextflow.Session session) {
// try using the old method
try {
Class<?> scriptParserClass = Class.forName('nextflow.script.ScriptParser')
return scriptParserClass.getDeclaredConstructor(nextflow.Session).newInstance(session)
} catch (ClassNotFoundException e) {
// else try with the new method
try {
Class<?> scriptLoaderFactoryClass = Class.forName('nextflow.script.ScriptLoaderFactory')
def createMethod = scriptLoaderFactoryClass.getDeclaredMethod('create', nextflow.Session)
return createMethod.invoke(null, session) // null because create is static
} catch (ClassNotFoundException | NoSuchMethodException | IllegalAccessException | java.lang.reflect.InvocationTargetException e2) {
// Handle the case where neither class is found
throw new Exception("Neither nextflow.script.ScriptParser nor nextflow.script.ScriptLoaderFactory could be found. Is this a compatible Nextflow version?", e2)
}
}
}
// defaults
meta["defaults"] = [
// key to be used to trace the process and determine output names

View File

@@ -48,10 +48,10 @@
"log": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.log.log`. File with logging information [default = stdout]",
"help_text": "Type: `file`, default: `$id.$key.log.log`. File with logging information [default = stdout]."
"description": "Type: `file`, default: `$id.$key.log`. File with logging information [default = stdout]",
"help_text": "Type: `file`, default: `$id.$key.log`. File with logging information [default = stdout]."
,
"default":"$id.$key.log.log"
"default":"$id.$key.log"
}
@@ -59,10 +59,10 @@
"error": {
"type":
"string",
"description": "Type: `file`, default: `$id.$key.error.error`. File with error information [default = stderr]",
"help_text": "Type: `file`, default: `$id.$key.error.error`. File with error information [default = stderr]."
"description": "Type: `file`, default: `$id.$key.error`. File with error information [default = stderr]",
"help_text": "Type: `file`, default: `$id.$key.error`. File with error information [default = stderr]."
,
"default":"$id.$key.error.error"
"default":"$id.$key.error"
}

View File

@@ -80,6 +80,9 @@ info:
- "modules/local/bedtools_genomecov.nf"
last_sha: "0a1bdcfbb498987643b74e9fccab85ccd9f2a17d"
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -182,15 +185,15 @@ build_info:
engine: "docker|native"
output: "target/executable/bedtools_genomecov"
executable: "target/executable/bedtools_genomecov/bedtools_genomecov"
viash_version: "0.9.0"
git_commit: "2b3d511b34246648b934fd1dc99b22e0a71c37f2"
git_remote: "https://x-access-token:ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
viash_version: "0.9.2"
git_commit: "f52978a0e25cae182b7874b4b8aa3afc183e880e"
git_remote: "https://github.com/viash-hub/rnaseq"
package_config:
name: "rnaseq"
version: "main"
info:
test_resources:
- path: "gs://viash-hub-test-data/rnaseq/v1"
- path: "gs://viash-hub-resources/rnaseq/v1"
dest: "testData"
repositories:
- type: "vsh"
@@ -201,7 +204,7 @@ package_config:
name: "craftbox"
repo: "craftbox"
tag: "v0.1.0"
viash_version: "0.9.0"
viash_version: "0.9.2"
source: "src"
target: "target"
config_mods:

View File

@@ -2,7 +2,7 @@
# bedtools_genomecov main
#
# This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
# This wrapper script is auto-generated by viash 0.9.2 and is thus a derivative
# work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
# Intuitive.
#
@@ -169,35 +169,6 @@ VIASH_META_CONFIG="$VIASH_META_RESOURCES_DIR/.config.vsh.yaml"
VIASH_META_TEMP_DIR="$VIASH_TEMP"
# ViashHelp: Display helpful explanation about this executable
function ViashHelp {
echo "bedtools_genomecov main"
echo ""
echo "Compute BEDGRAPH (-bg) summaries of feature coverage"
echo ""
echo "Input:"
echo " --strandedness"
echo " type: string"
echo " choices: [ unstranded, forward, reverse, auto ]"
echo " Sample strand-specificity."
echo ""
echo " --bam"
echo " type: file, file must exist"
echo " Genome BAM file"
echo ""
echo " --extra_bedtools_args"
echo " type: string"
echo " default:"
echo ""
echo "Output:"
echo " --bedgraph_forward"
echo " type: file, output, file must exist"
echo " default: \$id.forward.bedgraph"
echo ""
echo " --bedgraph_reverse"
echo " type: file, output, file must exist"
echo " default: \$id.reverse.bedgraph"
}
# initialise variables
VIASH_MODE='run'
@@ -481,9 +452,9 @@ mv bedtools.static /usr/local/bin/bedtools && \
chmod a+x /usr/local/bin/bedtools
LABEL org.opencontainers.image.description="Companion container for running component bedtools_genomecov"
LABEL org.opencontainers.image.created="2024-12-05T14:40:12Z"
LABEL org.opencontainers.image.source="https://x-access-token/ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
LABEL org.opencontainers.image.revision="2b3d511b34246648b934fd1dc99b22e0a71c37f2"
LABEL org.opencontainers.image.created="2025-05-06T19:18:33Z"
LABEL org.opencontainers.image.source="https://github.com/viash-hub/rnaseq"
LABEL org.opencontainers.image.revision="f52978a0e25cae182b7874b4b8aa3afc183e880e"
LABEL org.opencontainers.image.version="main"
VIASHDOCKER
@@ -598,6 +569,61 @@ fi
# initialise docker variables
VIASH_DOCKER_RUN_ARGS=(-i --rm)
# ViashHelp: Display helpful explanation about this executable
function ViashHelp {
echo "bedtools_genomecov main"
echo ""
echo "Compute BEDGRAPH (-bg) summaries of feature coverage"
echo ""
echo "Input:"
echo " --strandedness"
echo " type: string"
echo " choices: [ unstranded, forward, reverse, auto ]"
echo " Sample strand-specificity."
echo ""
echo " --bam"
echo " type: file, file must exist"
echo " Genome BAM file"
echo ""
echo " --extra_bedtools_args"
echo " type: string"
echo " default:"
echo ""
echo "Output:"
echo " --bedgraph_forward"
echo " type: file, output, file must exist"
echo " default: \$id.forward.bedgraph"
echo ""
echo " --bedgraph_reverse"
echo " type: file, output, file must exist"
echo " default: \$id.reverse.bedgraph"
echo ""
echo "Viash built in Computational Requirements:"
echo " ---cpus=INT"
echo " Number of CPUs to use"
echo " ---memory=STRING"
echo " Amount of memory to use. Examples: 4GB, 3MiB."
echo ""
echo "Viash built in Docker:"
echo " ---setup=STRATEGY"
echo " Setup the docker container. Options are: alwaysbuild, alwayscachedbuild, ifneedbebuild, ifneedbecachedbuild, alwayspull, alwayspullelsebuild, alwayspullelsecachedbuild, ifneedbepull, ifneedbepullelsebuild, ifneedbepullelsecachedbuild, push, pushifnotpresent, donothing."
echo " Default: ifneedbepullelsecachedbuild"
echo " ---dockerfile"
echo " Print the dockerfile to stdout."
echo " ---docker_run_args=ARG"
echo " Provide runtime arguments to Docker. See the documentation on \`docker run\` for more information."
echo " ---docker_image_id"
echo " Print the docker image id to stdout."
echo " ---debug"
echo " Enter the docker container for debugging purposes."
echo ""
echo "Viash built in Engines:"
echo " ---engine=ENGINE_ID"
echo " Specify the engine to use. Options are: docker, native."
echo " Default: docker"
}
# initialise array
VIASH_POSITIONAL_ARGS=''

View File

@@ -90,6 +90,9 @@ info:
- "modules/local/cat_additional_fasta.nf"
last_sha: "0a1bdcfbb498987643b74e9fccab85ccd9f2a17d"
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -186,15 +189,15 @@ build_info:
engine: "docker|native"
output: "target/executable/cat_additional_fasta"
executable: "target/executable/cat_additional_fasta/cat_additional_fasta"
viash_version: "0.9.0"
git_commit: "2b3d511b34246648b934fd1dc99b22e0a71c37f2"
git_remote: "https://x-access-token:ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
viash_version: "0.9.2"
git_commit: "f52978a0e25cae182b7874b4b8aa3afc183e880e"
git_remote: "https://github.com/viash-hub/rnaseq"
package_config:
name: "rnaseq"
version: "main"
info:
test_resources:
- path: "gs://viash-hub-test-data/rnaseq/v1"
- path: "gs://viash-hub-resources/rnaseq/v1"
dest: "testData"
repositories:
- type: "vsh"
@@ -205,7 +208,7 @@ package_config:
name: "craftbox"
repo: "craftbox"
tag: "v0.1.0"
viash_version: "0.9.0"
viash_version: "0.9.2"
source: "src"
target: "target"
config_mods:

View File

@@ -2,7 +2,7 @@
# cat_additional_fasta main
#
# This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
# This wrapper script is auto-generated by viash 0.9.2 and is thus a derivative
# work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
# Intuitive.
#
@@ -169,40 +169,6 @@ VIASH_META_CONFIG="$VIASH_META_RESOURCES_DIR/.config.vsh.yaml"
VIASH_META_TEMP_DIR="$VIASH_TEMP"
# ViashHelp: Display helpful explanation about this executable
function ViashHelp {
echo "cat_additional_fasta main"
echo ""
echo "Concatenate addional fasta file to reference FASTA and GTF files."
echo ""
echo "Input:"
echo " --fasta"
echo " type: file, required parameter, file must exist"
echo " Path to FASTA genome file."
echo ""
echo " --gtf"
echo " type: file, file must exist"
echo " Path to GTF annotation file."
echo ""
echo " --additional_fasta"
echo " type: file, file must exist"
echo " FASTA file to concatenate to genome FASTA file e.g. containing spike-in"
echo " sequences."
echo ""
echo " --biotype"
echo " type: string"
echo " Biotype value to use when appending entries to GTF file when additional"
echo " fasta file is provided."
echo ""
echo "Output:"
echo " --fasta_output"
echo " type: file, output, file must exist"
echo " Concatenated FASTA file."
echo ""
echo " --gtf_output"
echo " type: file, output, file must exist"
echo " Concatenated GTF file."
}
# initialise variables
VIASH_MODE='run'
@@ -480,9 +446,9 @@ function ViashDockerfile {
FROM python:latest
ENTRYPOINT []
LABEL org.opencontainers.image.description="Companion container for running component cat_additional_fasta"
LABEL org.opencontainers.image.created="2024-12-05T14:40:11Z"
LABEL org.opencontainers.image.source="https://x-access-token/ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
LABEL org.opencontainers.image.revision="2b3d511b34246648b934fd1dc99b22e0a71c37f2"
LABEL org.opencontainers.image.created="2025-05-06T19:18:34Z"
LABEL org.opencontainers.image.source="https://github.com/viash-hub/rnaseq"
LABEL org.opencontainers.image.revision="f52978a0e25cae182b7874b4b8aa3afc183e880e"
LABEL org.opencontainers.image.version="main"
VIASHDOCKER
@@ -597,6 +563,66 @@ fi
# initialise docker variables
VIASH_DOCKER_RUN_ARGS=(-i --rm)
# ViashHelp: Display helpful explanation about this executable
function ViashHelp {
echo "cat_additional_fasta main"
echo ""
echo "Concatenate addional fasta file to reference FASTA and GTF files."
echo ""
echo "Input:"
echo " --fasta"
echo " type: file, required parameter, file must exist"
echo " Path to FASTA genome file."
echo ""
echo " --gtf"
echo " type: file, file must exist"
echo " Path to GTF annotation file."
echo ""
echo " --additional_fasta"
echo " type: file, file must exist"
echo " FASTA file to concatenate to genome FASTA file e.g. containing spike-in"
echo " sequences."
echo ""
echo " --biotype"
echo " type: string"
echo " Biotype value to use when appending entries to GTF file when additional"
echo " fasta file is provided."
echo ""
echo "Output:"
echo " --fasta_output"
echo " type: file, output, file must exist"
echo " Concatenated FASTA file."
echo ""
echo " --gtf_output"
echo " type: file, output, file must exist"
echo " Concatenated GTF file."
echo ""
echo "Viash built in Computational Requirements:"
echo " ---cpus=INT"
echo " Number of CPUs to use"
echo " ---memory=STRING"
echo " Amount of memory to use. Examples: 4GB, 3MiB."
echo ""
echo "Viash built in Docker:"
echo " ---setup=STRATEGY"
echo " Setup the docker container. Options are: alwaysbuild, alwayscachedbuild, ifneedbebuild, ifneedbecachedbuild, alwayspull, alwayspullelsebuild, alwayspullelsecachedbuild, ifneedbepull, ifneedbepullelsebuild, ifneedbepullelsecachedbuild, push, pushifnotpresent, donothing."
echo " Default: ifneedbepullelsecachedbuild"
echo " ---dockerfile"
echo " Print the dockerfile to stdout."
echo " ---docker_run_args=ARG"
echo " Provide runtime arguments to Docker. See the documentation on \`docker run\` for more information."
echo " ---docker_image_id"
echo " Print the docker image id to stdout."
echo " ---debug"
echo " Enter the docker container for debugging purposes."
echo ""
echo "Viash built in Engines:"
echo " ---engine=ENGINE_ID"
echo " Specify the engine to use. Options are: docker, native."
echo " Default: docker"
}
# initialise array
VIASH_POSITIONAL_ARGS=''

View File

@@ -77,6 +77,9 @@ info:
- "modules/nf-core/cat/fastq/meta.yml"
last_sha: "54721c6946daf6d602d7069dc127deef9cbe6b33"
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -173,15 +176,15 @@ build_info:
engine: "docker|native"
output: "target/executable/cat_fastq"
executable: "target/executable/cat_fastq/cat_fastq"
viash_version: "0.9.0"
git_commit: "2b3d511b34246648b934fd1dc99b22e0a71c37f2"
git_remote: "https://x-access-token:ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
viash_version: "0.9.2"
git_commit: "f52978a0e25cae182b7874b4b8aa3afc183e880e"
git_remote: "https://github.com/viash-hub/rnaseq"
package_config:
name: "rnaseq"
version: "main"
info:
test_resources:
- path: "gs://viash-hub-test-data/rnaseq/v1"
- path: "gs://viash-hub-resources/rnaseq/v1"
dest: "testData"
repositories:
- type: "vsh"
@@ -192,7 +195,7 @@ package_config:
name: "craftbox"
repo: "craftbox"
tag: "v0.1.0"
viash_version: "0.9.0"
viash_version: "0.9.2"
source: "src"
target: "target"
config_mods:

View File

@@ -2,7 +2,7 @@
# cat_fastq main
#
# This wrapper script is auto-generated by viash 0.9.0 and is thus a derivative
# This wrapper script is auto-generated by viash 0.9.2 and is thus a derivative
# work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
# Intuitive.
#
@@ -169,32 +169,6 @@ VIASH_META_CONFIG="$VIASH_META_RESOURCES_DIR/.config.vsh.yaml"
VIASH_META_TEMP_DIR="$VIASH_TEMP"
# ViashHelp: Display helpful explanation about this executable
function ViashHelp {
echo "cat_fastq main"
echo ""
echo "Concatenate multiple fastq files"
echo ""
echo "Input:"
echo " --read_1"
echo " type: file, multiple values allowed, file must exist"
echo " Read 1 fastq files to be concatenated"
echo ""
echo " --read_2"
echo " type: file, multiple values allowed, file must exist"
echo " Read 2 fastq files to be concatenated"
echo ""
echo "Output:"
echo " --fastq_1"
echo " type: file, output, file must exist"
echo " default: \$id_r1.fastq"
echo " Concatenated read 1 fastq"
echo ""
echo " --fastq_2"
echo " type: file, output"
echo " default: \$id_r2.fastq"
echo " Concatenated read 2 fastq"
}
# initialise variables
VIASH_MODE='run'
@@ -472,9 +446,9 @@ function ViashDockerfile {
FROM ubuntu:22.04
ENTRYPOINT []
LABEL org.opencontainers.image.description="Companion container for running component cat_fastq"
LABEL org.opencontainers.image.created="2024-12-05T14:40:10Z"
LABEL org.opencontainers.image.source="https://x-access-token/ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
LABEL org.opencontainers.image.revision="2b3d511b34246648b934fd1dc99b22e0a71c37f2"
LABEL org.opencontainers.image.created="2025-05-06T19:18:31Z"
LABEL org.opencontainers.image.source="https://github.com/viash-hub/rnaseq"
LABEL org.opencontainers.image.revision="f52978a0e25cae182b7874b4b8aa3afc183e880e"
LABEL org.opencontainers.image.version="main"
VIASHDOCKER
@@ -589,6 +563,58 @@ fi
# initialise docker variables
VIASH_DOCKER_RUN_ARGS=(-i --rm)
# ViashHelp: Display helpful explanation about this executable
function ViashHelp {
echo "cat_fastq main"
echo ""
echo "Concatenate multiple fastq files"
echo ""
echo "Input:"
echo " --read_1"
echo " type: file, multiple values allowed, file must exist"
echo " Read 1 fastq files to be concatenated"
echo ""
echo " --read_2"
echo " type: file, multiple values allowed, file must exist"
echo " Read 2 fastq files to be concatenated"
echo ""
echo "Output:"
echo " --fastq_1"
echo " type: file, output, file must exist"
echo " default: \$id_r1.fastq"
echo " Concatenated read 1 fastq"
echo ""
echo " --fastq_2"
echo " type: file, output"
echo " default: \$id_r2.fastq"
echo " Concatenated read 2 fastq"
echo ""
echo "Viash built in Computational Requirements:"
echo " ---cpus=INT"
echo " Number of CPUs to use"
echo " ---memory=STRING"
echo " Amount of memory to use. Examples: 4GB, 3MiB."
echo ""
echo "Viash built in Docker:"
echo " ---setup=STRATEGY"
echo " Setup the docker container. Options are: alwaysbuild, alwayscachedbuild, ifneedbebuild, ifneedbecachedbuild, alwayspull, alwayspullelsebuild, alwayspullelsecachedbuild, ifneedbepull, ifneedbepullelsebuild, ifneedbepullelsecachedbuild, push, pushifnotpresent, donothing."
echo " Default: ifneedbepullelsecachedbuild"
echo " ---dockerfile"
echo " Print the dockerfile to stdout."
echo " ---docker_run_args=ARG"
echo " Provide runtime arguments to Docker. See the documentation on \`docker run\` for more information."
echo " ---docker_image_id"
echo " Print the docker image id to stdout."
echo " ---debug"
echo " Enter the docker container for debugging purposes."
echo ""
echo "Viash built in Engines:"
echo " ---engine=ENGINE_ID"
echo " Specify the engine to use. Options are: docker, native."
echo " Default: docker"
}
# initialise array
VIASH_POSITIONAL_ARGS=''

View File

@@ -0,0 +1,183 @@
name: "copy_if_exists"
version: "main"
argument_groups:
- name: "Input"
arguments:
- type: "file"
name: "--required_file"
info: null
example:
- "/tmp/rnaseq_workflow_config/required_file.txt"
must_exist: false
create_parent: true
required: true
direction: "input"
multiple: false
multiple_sep: ";"
- type: "file"
name: "--optional_file"
info: null
example:
- "/tmp/rnaseq_workflow_config/optional_file.txt"
must_exist: false
create_parent: true
required: false
direction: "input"
multiple: false
multiple_sep: ";"
- name: "Ouput"
arguments:
- type: "file"
name: "--output"
info: null
default:
- "copy_if_exists_output"
must_exist: true
create_parent: true
required: false
direction: "output"
multiple: false
multiple_sep: ";"
resources:
- type: "bash_script"
path: "script.sh"
is_executable: true
- type: "file"
path: "required_file.txt"
- type: "file"
path: "optional_file.txt"
- type: "file"
path: "nextflow_labels.config"
dest: "nextflow_labels.config"
info: null
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
repositories:
- type: "vsh"
name: "biobox"
repo: "biobox"
tag: "main"
- type: "vsh"
name: "craftbox"
repo: "craftbox"
tag: "v0.1.0"
runners:
- type: "executable"
id: "executable"
docker_setup_strategy: "ifneedbepullelsecachedbuild"
- type: "nextflow"
id: "nextflow"
directives:
tag: "$id"
auto:
simplifyInput: true
simplifyOutput: false
transcript: false
publish: false
config:
labels:
mem1gb: "memory = 1000000000.B"
mem2gb: "memory = 2000000000.B"
mem5gb: "memory = 5000000000.B"
mem10gb: "memory = 10000000000.B"
mem20gb: "memory = 20000000000.B"
mem50gb: "memory = 50000000000.B"
mem100gb: "memory = 100000000000.B"
mem200gb: "memory = 200000000000.B"
mem500gb: "memory = 500000000000.B"
mem1tb: "memory = 1000000000000.B"
mem2tb: "memory = 2000000000000.B"
mem5tb: "memory = 5000000000000.B"
mem10tb: "memory = 10000000000000.B"
mem20tb: "memory = 20000000000000.B"
mem50tb: "memory = 50000000000000.B"
mem100tb: "memory = 100000000000000.B"
mem200tb: "memory = 200000000000000.B"
mem500tb: "memory = 500000000000000.B"
mem1gib: "memory = 1073741824.B"
mem2gib: "memory = 2147483648.B"
mem4gib: "memory = 4294967296.B"
mem8gib: "memory = 8589934592.B"
mem16gib: "memory = 17179869184.B"
mem32gib: "memory = 34359738368.B"
mem64gib: "memory = 68719476736.B"
mem128gib: "memory = 137438953472.B"
mem256gib: "memory = 274877906944.B"
mem512gib: "memory = 549755813888.B"
mem1tib: "memory = 1099511627776.B"
mem2tib: "memory = 2199023255552.B"
mem4tib: "memory = 4398046511104.B"
mem8tib: "memory = 8796093022208.B"
mem16tib: "memory = 17592186044416.B"
mem32tib: "memory = 35184372088832.B"
mem64tib: "memory = 70368744177664.B"
mem128tib: "memory = 140737488355328.B"
mem256tib: "memory = 281474976710656.B"
mem512tib: "memory = 562949953421312.B"
cpu1: "cpus = 1"
cpu2: "cpus = 2"
cpu5: "cpus = 5"
cpu10: "cpus = 10"
cpu20: "cpus = 20"
cpu50: "cpus = 50"
cpu100: "cpus = 100"
cpu200: "cpus = 200"
cpu500: "cpus = 500"
cpu1000: "cpus = 1000"
script:
- "includeConfig(\"nextflow_labels.config\")"
debug: false
container: "docker"
engines:
- type: "docker"
id: "docker"
image: "ubuntu:22.04"
target_registry: "images.viash-hub.com"
target_tag: "main"
namespace_separator: "/"
entrypoint: []
cmd: null
- type: "native"
id: "native"
build_info:
config: "src/extra/copy_if_exists/config.vsh.yaml"
runner: "executable"
engine: "docker|native"
output: "target/executable/copy_if_exists"
executable: "target/executable/copy_if_exists/copy_if_exists"
viash_version: "0.9.2"
git_commit: "f52978a0e25cae182b7874b4b8aa3afc183e880e"
git_remote: "https://github.com/viash-hub/rnaseq"
package_config:
name: "rnaseq"
version: "main"
info:
test_resources:
- path: "gs://viash-hub-resources/rnaseq/v1"
dest: "testData"
repositories:
- type: "vsh"
name: "biobox"
repo: "biobox"
tag: "main"
- type: "vsh"
name: "craftbox"
repo: "craftbox"
tag: "v0.1.0"
viash_version: "0.9.2"
source: "src"
target: "target"
config_mods:
- ".requirements.commands := ['ps']\n.resources += {path: '/src/workflows/utils/labels.config',\
\ dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].directives.tag\
\ := '$id'\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\"nextflow_labels.config\"\
)'\n"
- ".engines += { type: \"native\" }"
- ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
- ".engines[.type == 'docker'].target_tag := 'main'"
organization: "vsh"

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,44 @@
process {
// Default resources for components that hardly do any processing
memory = { 2.GB * task.attempt }
cpus = 1
// Retry for exit codes that have something to do with memory issues
errorStrategy = { task.exitStatus in 137..140 ? 'retry' : 'terminate' }
maxRetries = 3
maxMemory = null
// Resource labels
withLabel: singlecpu { cpus = 1 }
withLabel: lowcpu { cpus = 2 }
withLabel: midcpu { cpus = 4 }
withLabel: highcpu { cpus = 8 }
withLabel: veryhighcpu { cpus = 16 }
withLabel: lowmem { memory = { get_memory( 4.GB * task.attempt ) } }
withLabel: midmem { memory = { get_memory( 16.GB * task.attempt ) } }
withLabel: highmem { memory = { get_memory( 24.GB * task.attempt ) } }
withLabel: veryhighmem { memory = { get_memory( 48.GB * task.attempt ) } }
}
def get_memory(to_compare) {
if (!process.containsKey("maxMemory") || !process.maxMemory) {
return to_compare
}
try {
if (process.containsKey("maxRetries") && process.maxRetries && task.attempt == (process.maxRetries as int)) {
return process.maxMemory
}
else if (to_compare.compareTo(process.maxMemory as nextflow.util.MemoryUnit) == 1) {
return max_memory as nextflow.util.MemoryUnit
}
else {
return to_compare
}
} catch (all) {
println "Error processing memory resources. Please check that process.maxMemory '${process.maxMemory}' and process.maxRetries '${process.maxRetries}' are valid!"
System.exit(1)
}
}

View File

@@ -0,0 +1 @@
Optional!

View File

@@ -0,0 +1 @@
Required!

View File

@@ -133,6 +133,9 @@ info:
- "modules/local/deseq2_qc.nf"
last_sha: "92b2a7857de1dda9d1c19a088941fc81e2976ff7"
status: "enabled"
scope:
image: "public"
target: "public"
requirements:
commands:
- "ps"
@@ -215,11 +218,20 @@ runners:
engines:
- type: "docker"
id: "docker"
image: "rocker/r2u:22.04"
image: "debian:latest"
target_registry: "images.viash-hub.com"
target_tag: "main"
namespace_separator: "/"
setup:
- type: "apt"
packages:
- "libcurl4-openssl-dev"
- "r-base"
- "r-base-core"
- "libxml2-dev"
- "procps"
- "libssl-dev"
interactive: false
- type: "r"
cran:
- "optparse"
@@ -231,6 +243,7 @@ engines:
bioc:
- "DESeq2"
bioc_force_install: false
warnings_as_errors: true
entrypoint: []
cmd: null
- type: "native"
@@ -241,15 +254,15 @@ build_info:
engine: "docker|native"
output: "target/executable/deseq2_qc"
executable: "target/executable/deseq2_qc/deseq2_qc"
viash_version: "0.9.0"
git_commit: "2b3d511b34246648b934fd1dc99b22e0a71c37f2"
git_remote: "https://x-access-token:ghs_mpjpezoXRPhvITbEzvFPaxzcp8yfML2ITu9P@github.com/viash-hub/rnaseq"
viash_version: "0.9.2"
git_commit: "f52978a0e25cae182b7874b4b8aa3afc183e880e"
git_remote: "https://github.com/viash-hub/rnaseq"
package_config:
name: "rnaseq"
version: "main"
info:
test_resources:
- path: "gs://viash-hub-test-data/rnaseq/v1"
- path: "gs://viash-hub-resources/rnaseq/v1"
dest: "testData"
repositories:
- type: "vsh"
@@ -260,7 +273,7 @@ package_config:
name: "craftbox"
repo: "craftbox"
tag: "v0.1.0"
viash_version: "0.9.0"
viash_version: "0.9.2"
source: "src"
target: "target"
config_mods:

Some files were not shown because too many files have changed in this diff Show More