SamStudio8 · July 28, 2021 17:00
diff --git a/collect.nf b/collect.nf
 #!/usr/bin/env nextflow

 params.testdir = "/cephfs/covid/software/sam/nf-test/"

 params.fofn_single = [params.testdir, 'samples.csv'].join('/')
 single_manifest_ch = Channel
    .fromPath(params.fofn_single)  // open say, a CSV
    .splitCsv(header:true)         // split text stream into CSV records
    .map { row-> file(row.path) }  // coerce the record and return a file for each line
    .collect()                     // emit all items as one

 process echo_fofn_single {

    input:
    file '*.txt' from single_manifest_ch

    output:
    publishDir path: "${params.testdir}", pattern: "out", mode: "copy", overwrite: true
    file "out"

    script:
    """
    cat *.txt > out
    """
 }


 params.fofn_group = [params.testdir, 'grouped_samples.tsv'].join('/')
 group_manifest_ch = Channel
    .fromPath(params.fofn_group)
    .splitCsv(header:true, sep:'\t')
    .map { row-> tuple(row.group, file(row.path)) }  // coerce the record into a (group, file) tuple
    .groupTuple()                                    // return tuples of (group, [file1, ...fileN])

 process echo_fofn_group {

    input:
    tuple(group, file(files)) from group_manifest_ch

    output:
    publishDir path: "${params.testdir}", pattern: "g*out", mode: "copy", overwrite: true
    file "g${group}.out"

    script:
    """
    cat ${files} > g${group}.out
    """
 }

diff --git a/file1.txt b/file1.txt
 hello
diff --git a/file2.txt b/file2.txt
 world
diff --git a/file3.txt b/file3.txt
 hoot
diff --git a/file4.txt b/file4.txt
 meow
diff --git a/file5.txt b/file5.txt
 bang
diff --git a/g1.out b/g1.out
 hello
 world
diff --git a/g2.out b/g2.out
 hoot
 meow
diff --git a/g3.out b/g3.out
 bang
diff --git a/grouped_samples.tsv b/grouped_samples.tsv
diff --git a/out b/out
 hello
 world
diff --git a/samples.csv b/samples.csv
 path
 /cephfs/covid/software/sam/nf-test/file1.txt
 /cephfs/covid/software/sam/nf-test/file2.txt
	#!/usr/bin/env nextflow

	params.testdir = "/cephfs/covid/software/sam/nf-test/"

	params.fofn_single = [params.testdir, 'samples.csv'].join('/')
	single_manifest_ch = Channel
	.fromPath(params.fofn_single) // open say, a CSV
	.splitCsv(header:true) // split text stream into CSV records
	.map { row-> file(row.path) } // coerce the record and return a file for each line
	.collect() // emit all items as one

	process echo_fofn_single {

	input:
	file '*.txt' from single_manifest_ch

	output:
	publishDir path: "${params.testdir}", pattern: "out", mode: "copy", overwrite: true
	file "out"

	script:
	"""
	cat *.txt > out
	"""
	}


	params.fofn_group = [params.testdir, 'grouped_samples.tsv'].join('/')
	group_manifest_ch = Channel
	.fromPath(params.fofn_group)
	.splitCsv(header:true, sep:'\t')
	.map { row-> tuple(row.group, file(row.path)) } // coerce the record into a (group, file) tuple
	.groupTuple() // return tuples of (group, [file1, ...fileN])

	process echo_fofn_group {

	input:
	tuple(group, file(files)) from group_manifest_ch

	output:
	publishDir path: "${params.testdir}", pattern: "g*out", mode: "copy", overwrite: true
	file "g${group}.out"

	script:
	"""
	cat ${files} > g${group}.out
	"""
	}
	group	path
	1	/cephfs/covid/software/sam/nf-test/file1.txt
	1	/cephfs/covid/software/sam/nf-test/file2.txt
	2	/cephfs/covid/software/sam/nf-test/file3.txt
	2	/cephfs/covid/software/sam/nf-test/file4.txt
	3	/cephfs/covid/software/sam/nf-test/file5.txt