add crabs/dbimport from readsimulator pipeline and rename to crabs/import

modules/nf-core/crabs/import/environment.yml

@@ -0,0 +1,5 @@
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - bioconda::crabs=1.0.7

modules/nf-core/crabs/import/main.nf

@@ -0,0 +1,59 @@
+process CRABS_IMPORT {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/crabs:1.0.7--pyhdfd78af_0':
+        'biocontainers/crabs:1.0.7--pyhdfd78af_0' }"
+
+    input:
+    tuple val(meta), path(fasta)
+    tuple val(meta2), path(accession2taxid)
+    tuple val(meta3), path(names)
+    tuple val(meta4), path(nodes)
+
+    output:
+    tuple val(meta), path("*.fa"), emit: fasta
+    path "versions.yml"          , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args          = task.ext.args ?: ''
+    def prefix        = task.ext.prefix ?: "${meta.id}"
+    def is_compressed = fasta.name.endsWith(".gz")
+    def fasta_name    = fasta.name.replace(".gz", "")
+    """
+    if [ "${is_compressed}" == "true" ]; then
+        gzip -c -d ${fasta} > ${fasta_name}
+    fi
+
+    crabs --import \\
+        --input ${fasta_name} \\
+        --output ${prefix}.crabsdb.fa \\
+        --acc2tax ${accession2taxid} \\
+        --names ${names} \\
+        --nodes ${nodes} \\
+        $args
+
+    rm ${fasta_name}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        crabs: \$(crabs --help 2>/dev/null | grep 'CRABS |' | sed 's/.*CRABS | v\\([0-9.]*\\).*/\\1/')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}.fa
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        crabs: \$(crabs --help 2>/dev/null | grep 'CRABS |' | sed 's/.*CRABS | v\\([0-9.]*\\).*/\\1/')
+    END_VERSIONS
+    """
+}

modules/nf-core/crabs/import/meta.yml

@@ -0,0 +1,79 @@
+name: crabs_import
+description: In-house generated or curated data can be imported into CRABS.
+keywords:
+  - insilico
+  - amplicon
+  - sequencing
+  - inhouse
+tools:
+  - crabs:
+      description: |
+        Crabs (Creating Reference databases for Amplicon-Based Sequencing)
+        is a program to download and curate reference databases
+        for eDNA metabarcoding analyses
+      homepage: https://github.com/gjeunen/reference_database_creator
+      documentation: https://github.com/gjeunen/reference_database_creator?tab=readme-ov-file#running-crabs
+      tool_dev_url: https://github.com/gjeunen/reference_database_creator
+      doi: 10.1111/1755-0998.13741
+      licence: ["MIT License"]
+      identifier: ""
+
+input:
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1' ]`
+    - fasta:
+        type: file
+        description: In-house sequencing data
+        pattern: "*.{fa,fasta}"
+  - - meta2:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1' ]`
+    - accession2taxid:
+        type: file
+        description: Linking accession numbers to taxonomic IDs
+        pattern: "*.accession2taxid"
+  - - meta3:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1' ]`
+    - names:
+        type: file
+        description: Containing information about the phylogenetic name associated with each taxonomic ID 
+        pattern: "*.dmp"
+  - - meta4: 
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1' ]`
+    - nodes:
+        type: file
+        description: Containing information how taxonomic IDs are linked
+        pattern: "*.dmp"
+output:
+  - fasta:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*.fa":
+          type: file
+          description: Reverse complemented Sequence
+          pattern: "*.{fa,fasta}"
+  - versions:
+      - versions.yml:
+          type: file
+          description: File containing software versions
+          pattern: "versions.yml"
+authors:
+  - "@famosab"
+  - "@a4000"
+maintainers:
+  - "@famosab"
+  - "@a4000"

modules/nf-core/crabs/import/tests/main.nf.test

@@ -0,0 +1,90 @@
+nextflow_process {
+
+    name "Test Process CRABS_IMPORT"
+    script "../main.nf"
+    process "CRABS_IMPORT"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "crabs"
+    tag "crabs/import"
+
+    test("sarscov2 - fasta") {
+
+        config "./nextflow.config"
+
+        when {
+            params {
+                module_args = "--import-format embl --ranks 'superkingdom;phylum;class;order;family;genus;species' \\"
+            }
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/sarscov2/genome/genome-ena.fasta', checkIfExists: true)
+                ]
+                input[1] = [
+                    [ id:'accession2taxid' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/prokaryotes/metagenome/taxonomy/accession2taxid/nucl_gb.accession2taxid', checkIfExists: true)
+                ]
+                input[2] = [
+                    [ id:'names' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/prokaryotes/metagenome/taxonomy/taxdmp/names.dmp', checkIfExists: true)
+                ]
+                input[3] = [
+                    [ id:'nodes' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/prokaryotes/metagenome/taxonomy/taxdmp/nodes.dmp', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - fasta - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/sarscov2/genome/genome-ena.fasta', checkIfExists: true)
+                ]
+                input[1] = [
+                    [ id:'accession2taxid' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/prokaryotes/metagenome/taxonomy/accession2taxid/nucl_gb.accession2taxid', checkIfExists: true)
+                ]
+                input[2] = [
+                    [ id:'names' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/prokaryotes/metagenome/taxonomy/taxdmp/names.dmp', checkIfExists: true)
+                ]
+                input[3] = [
+                    [ id:'nodes' ], // meta map
+                    file(params.modules_testdata_base_path +'genomics/prokaryotes/metagenome/taxonomy/taxdmp/nodes.dmp', checkIfExists: true)
+                ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(
+                    process.out,
+                    path(process.out.versions[0]).yaml
+                    ).match() }
+            )
+        }
+
+    }
+
+}

modules/nf-core/crabs/import/tests/main.nf.test.snap

@@ -0,0 +1,73 @@
+{
+    "sarscov2 - fasta - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.fa:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,d21925bd41f67039425ac09278c847d0"
+                ],
+                "fasta": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.fa:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,d21925bd41f67039425ac09278c847d0"
+                ]
+            },
+            {
+                "CRABS_IMPORT": {
+                    "crabs": "1.0.7"
+                }
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.4"
+        },
+        "timestamp": "2025-02-19T13:58:38.53029566"
+    },
+    "sarscov2 - fasta": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.crabsdb.fa:md5,22c8c613e67ce3f5237a6c4646a0abb1"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,d21925bd41f67039425ac09278c847d0"
+                ],
+                "fasta": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.crabsdb.fa:md5,22c8c613e67ce3f5237a6c4646a0abb1"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,d21925bd41f67039425ac09278c847d0"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.4"
+        },
+        "timestamp": "2025-02-19T13:58:30.649755841"
+    }
+}

modules/nf-core/crabs/import/tests/nextflow.config

@@ -0,0 +1,7 @@
+process {
+
+    withName: CRABS_IMPORT {
+        ext.args = params.module_args
+    }
+
+}