Update md5sum module to be able to take multiple files

modules/nf-core/md5sum/main.nf

@@ -8,7 +8,8 @@ process MD5SUM {
         'nf-core/ubuntu:20.04' }"
 
     input:
-    tuple val(meta), path(file)
+    tuple val(meta), path(files)
+    val as_separate_files
 
     output:
     tuple val(meta), path("*.md5"), emit: checksum
@@ -19,29 +20,55 @@ process MD5SUM {
 
     script:
     def args = task.ext.args ?: ''
-    def prefix = task.ext.prefix ?: "${meta.id}"
+    def prefix = task.ext.prefix ?: "${meta.id}" // will only use when as_separate_files = false
+    if ( as_separate_files ) {
+        """
+        find -L * -type f \\
+            ! -name '*.md5' \\
+            -exec sh -c 'md5sum $args "\$1" > "\$1.md5"' _ "{}" \\;
 
-    """
-    md5sum \\
-        $args \\
-        ${file} \\
-        > ${file}.md5
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            md5sum: \$(echo \$(md5sum --version 2>&1 | head -n 1| sed 's/^.*) //;' ))
+        END_VERSIONS
+        """
+    } else {
+        """
+        find -L * -type f \\
+            ! -name '*.md5' \\
+            -exec md5sum $args "{}" + \\
+            > ${prefix}.md5
 
-    cat <<-END_VERSIONS > versions.yml
-    "${task.process}":
-        md5sum: \$(echo \$(md5sum --version 2>&1 | head -n 1| sed 's/^.*) //;' ))
-    END_VERSIONS
-    """
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            md5sum: \$(echo \$(md5sum --version 2>&1 | head -n 1| sed 's/^.*) //;' ))
+        END_VERSIONS
+        """
+    }
 
     stub:
     def args = task.ext.args ?: ''
-    """
-    touch ${file}.md5
-
-    cat <<-END_VERSIONS > versions.yml
-    "${task.process}":
-        md5sum: \$(echo \$(md5sum --version 2>&1 | head -n 1| sed 's/^.*) //;' ))
-    END_VERSIONS
-    """
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    if ( as_separate_files ) {
+        """
+        find -L * -type f \\
+            ! -name '*.md5' \\
+            -exec sh -c 'touch "\$1.md5"' _ "{}" \\;
+
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            md5sum: \$(echo \$(md5sum --version 2>&1 | head -n 1| sed 's/^.*) //;' ))
+        END_VERSIONS
+        """
+    } else {
+        """
+        touch ${prefix}.md5
+
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            md5sum: \$(echo \$(md5sum --version 2>&1 | head -n 1| sed 's/^.*) //;' ))
+        END_VERSIONS
+        """
+    }
 
 }

modules/nf-core/md5sum/meta.yml

@@ -1,12 +1,12 @@
 name: "md5sum"
-description: Create an MD5 (128-bit) checksum
+description: Create MD5 (128-bit) checksums
 keywords:
   - checksum
   - MD5
   - 128 bit
 tools:
   - "md5sum":
-      description: Create an MD5 (128-bit) checksum
+      description: Create MD5 (128-bit) checksums for each file
       homepage: "https://www.gnu.org"
       documentation: "https://man7.org/linux/man-pages/man1/md5sum.1.html"
       licence: ["GPL-3.0-or-later"]
@@ -16,9 +16,9 @@ input:
       description: |
         Groovy Map containing sample information
         e.g. [ id:'test', single_end:false ]
-  - file:
+  - files:
       type: file
-      description: Any file
+      description: Any number of files. One md5sum file will be generated for each.
       pattern: "*.*"
 output:
   - meta:

modules/nf-core/md5sum/tests/main.nf.test

@@ -17,6 +17,7 @@ nextflow_process {
                     [ id: 'hello' ],
                     [ file(params.modules_testdata_base_path + 'generic/txt/hello.txt', checkIfExists: true) ]
                 ]
+                input[1] = true
                 """
             }
         }
@@ -40,6 +41,7 @@ nextflow_process {
                     [ id: 'hello' ],
                     [ file(params.modules_testdata_base_path + 'generic/txt/hello.txt', checkIfExists: true) ]
                 ]
+                input[1] = true
                 """
             }
         }
@@ -51,5 +53,128 @@ nextflow_process {
             }
         }
     }
-
+
+    test("md5sum on hello.txt - stub") {
+        options "-stub"
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id: 'hello' ],
+                    [ file(params.modules_testdata_base_path + 'generic/txt/hello.txt', checkIfExists: true) ]
+                ]
+                input[1] = true
+                """
+            }
+        }
+
+        then {
+            assertAll {
+                { assert process.success }
+                { assert snapshot(process.out).match() }
+            }
+        }
+    }
+
+    test("md5sum on paired fastq, combined") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id: 'test' ],
+                    [
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_1.fastq.gz', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_2.fastq.gz', checkIfExists: true)
+                    ]
+                ]
+                input[1] = false
+                """
+            }
+        }
+
+        then {
+            assertAll {
+                { assert process.success }
+                { assert snapshot(process.out).match() }
+            }
+        }
+    }
+
+    test("md5sum on paired fastq, combined - stub") {
+        options "-stub"
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id: 'test' ],
+                    [
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_1.fastq.gz', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_2.fastq.gz', checkIfExists: true)
+                    ]
+                ]
+                input[1] = false
+                """
+            }
+        }
+
+        then {
+            assertAll {
+                { assert process.success }
+                { assert snapshot(process.out).match() }
+            }
+        }
+    }
+
+
+    test("md5sum on paired fastq, separate") {
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id: 'test' ],
+                    [
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_1.fastq.gz', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_2.fastq.gz', checkIfExists: true)
+                    ]
+                ]
+                input[1] = true
+                """
+            }
+        }
+
+        then {
+            assertAll {
+                { assert process.success }
+                { assert snapshot(process.out).match() }
+            }
+        }
+    }
+
+    test("md5sum on paired fastq, separate - stub") {
+        options "-stub"
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id: 'test' ],
+                    [
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_1.fastq.gz', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test_2.fastq.gz', checkIfExists: true)
+                    ]
+                ]
+                input[1] = true
+                """
+            }
+        }
+
+        then {
+            assertAll {
+                { assert process.success }
+                { assert snapshot(process.out).match() }
+            }
+        }
+    }
+
 }