Merge pull request #1678 from maxulysse/nft-utils-0-0-3

Migrate pytest aligner and pipeline default tests to nf-test

.github/workflows/ci.yml

@@ -312,6 +312,16 @@ jobs:
           mkdir -p $NXF_SINGULARITY_CACHEDIR
           mkdir -p $NXF_SINGULARITY_LIBRARYDIR
 
+      - name: Cache pdiff
+        uses: actions/cache@0c45773b623bea8c8e75f6c82b208c3cf94ea4f9 # v4
+        id: cache-pip-pdiff
+        with:
+          path: ~/.cache/pip
+          key: ${{ runner.os }}-pip-pdiff
+
+      - name: Install pdiff
+        run: python -m pip install --upgrade pip pdiff cryptography
+
       - name: Set up miniconda
         if: matrix.profile == 'conda'
         uses: conda-incubator/setup-miniconda@a4260408e20b96e80095f42ff7f1a15b27dd94ca # v3

CHANGELOG.md

@@ -18,6 +18,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Changed
 
 - [1669](https://github.com/nf-core/sarek/pull/1669) - Better nf-test pipeline level tests
+- [1677](https://github.com/nf-core/sarek/pull/1677) - Migrate pytest aligner and pipeline default tests to nf-test
 - [1680](https://github.com/nf-core/sarek/pull/1680) - Template update for nf-core/tools v3.0.0
 - [1681](https://github.com/nf-core/sarek/pull/1681) - Template update for nf-core/tools v3.0.1
 

nf-test.config

@@ -1,10 +1,19 @@
 config {
+    // location for all nf-tests
     testsDir "."
-    workDir ".nf-test"
+
+    // nf-test directory including temporary files for each test
+    workDir System.getenv("NFT_WORKDIR") ?: ".nf-test"
+
+    // location of an optional nextflow.config file specific for executing tests
     configFile "conf/test.config"
+
+    // run all test with defined profile(s) from the main nextflow.config
     profile "test"
 
+    // Include plugins
     plugins {
-        load "[email protected]"
+        load "[email protected]"
+        load "[email protected]"
     }
 }

tests/.nftignore

@@ -1,4 +1,5 @@
 csv/*.csv
+multiqc/multiqc_data/gatk-base-recalibrator-reported-empirical-plot.txt
 multiqc/multiqc_data/gatk_base_recalibrator.txt
 multiqc/multiqc_data/multiqc.log
 multiqc/multiqc_data/multiqc_data.json
@@ -7,12 +8,12 @@ multiqc/multiqc_data/multiqc_picard_dups.txt
 multiqc/multiqc_data/multiqc_software_versions.txt
 multiqc/multiqc_data/multiqc_sources.txt
 multiqc/multiqc_data/picard_deduplication.txt
-multiqc/multiqc_plots/pdf/*
-multiqc/multiqc_plots/png/picard_deduplication*.png
-multiqc/multiqc_plots/svg/*
+multiqc/multiqc_plots/{svg,pdf,png}/*.{svg,pdf,png}
 multiqc/multiqc_report.html
 pipeline_info/*.{html,json,txt,yml}
-preprocessing/**/*.{md,recal}.{cram,cram.crai,table}
+preprocessing/**/*.{md,recal,sorted}.{bam,bam.bai,cram,cram.crai,table}
+reference/dragmap/hash_table.{cfg,cfg.bin}
+reference/dragmap/hash_table_stats.txt
 reports/fastqc/**/*_fastqc.{html,zip}
 reports/markduplicates/**/*.md.cram.metrics
 variant_calling/**/*.vcf.{gz,gz.tbi}

tests/aligner-bwa-mem.nf.test

@@ -0,0 +1,81 @@
+nextflow_pipeline {
+
+    name "Test pipeline"
+    script "../main.nf"
+    tag "pipeline"
+    tag "pipeline_sarek"
+
+    test("Run with profile test | --aligner bwa-mem --save_reference | skip QC/recal/md") {
+
+        when {
+            params {
+                aligner = 'bwa-mem'
+                modules_testdata_base_path = 'https://raw.githubusercontent.com/nf-core/test-datasets/modules/data/'
+                outdir = "$outputDir"
+                save_reference = true
+                skip_tools = 'baserecalibrator,fastqc,markduplicates,mosdepth,multiqc,samtools'
+                tools = ''
+            }
+        }
+
+        then {
+            // stable_name: All files + folders in ${params.outdir}/ with a stable name
+            def stable_name = getAllFilesFromDir(params.outdir, relative: true, includeDir: true, ignore: ['pipeline_info/*.{html,json,txt}'])
+            // stable_path: All files in ${params.outdir}/ with stable content
+            def stable_path = getAllFilesFromDir(params.outdir, ignoreFile: 'tests/.nftignore')
+            // cram_files: All cram files
+            def cram_files  = getAllFilesFromDir(params.outdir, include: ['**/*.cram'])
+            def fasta       = params.modules_testdata_base_path + 'genomics/homo_sapiens/genome/genome.fasta'
+            assertAll(
+                { assert workflow.success},
+                { assert snapshot(
+                    // Number of successful tasks
+                    workflow.trace.succeeded().size(),
+                    // pipeline versions.yml file for multiqc from which Nextflow version is removed because we tests pipelines on multiple Nextflow versions
+                    removeNextflowVersion("$outputDir/pipeline_info/nf_core_sarek_software_mqc_versions.yml"),
+                    // All stable path name, with a relative path
+                    stable_name,
+                    // All files with stable contents
+                    stable_path,
+                    // All cram files
+                    cram_files.collect{ file -> [ file.getName(), cram(file.toString(), fasta).getReadsMD5() ] }
+                ).match() }
+            )
+        }
+    }
+
+    test("Run with profile test | --aligner bwa-mem --save_reference --build_only_index") {
+
+        when {
+            params {
+                aligner = 'bwa-mem'
+                build_only_index = true
+                input = false
+                outdir = "$outputDir"
+                save_reference = true
+                skip_tools = 'multiqc'
+                tools = ''
+            }
+        }
+
+        then {
+            // stable_name: All files + folders in ${params.outdir}/ with a stable name
+            def stable_name = getAllFilesFromDir(params.outdir, relative: true, includeDir: true, ignore: ['pipeline_info/*.{html,json,txt}'])
+            // stable_path: All files in ${params.outdir}/ with stable content
+            def stable_path = getAllFilesFromDir(params.outdir, ignoreFile: 'tests/.nftignore')
+            assertAll(
+                { assert workflow.success},
+                { assert snapshot(
+                    // Number of successful tasks
+                    workflow.trace.succeeded().size(),
+                    // pipeline versions.yml file for multiqc from which Nextflow version is removed because we tests pipelines on multiple Nextflow versions
+                    removeNextflowVersion("$outputDir/pipeline_info/nf_core_sarek_software_mqc_versions.yml"),
+                    // All stable path name, with a relative path
+                    stable_name,
+                    // All files with stable contents
+                    stable_path
+                ).match() }
+            )
+        }
+    }
+}

tests/aligner-bwa-mem.nf.test.snap

@@ -0,0 +1,113 @@
+{
+    "Run with profile test | --aligner bwa-mem --save_reference | skip QC/recal/md": {
+        "content": [
+            10,
+            {
+                "BAM_TO_CRAM_MAPPING": {
+                    "samtools": 1.21
+                },
+                "BWAMEM1_MEM": {
+                    "bwa": "0.7.18-r1243-dirty",
+                    "samtools": 1.2
+                },
+                "INDEX_MERGE_BAM": {
+                    "samtools": 1.21
+                },
+                "MERGE_BAM": {
+                    "samtools": 1.21
+                },
+                "Workflow": {
+                    "nf-core/sarek": "v3.5.0dev"
+                }
+            },
+            [
+                "csv",
+                "csv/mapped.csv",
+                "pipeline_info",
+                "pipeline_info/nf_core_sarek_software_mqc_versions.yml",
+                "preprocessing",
+                "preprocessing/mapped",
+                "preprocessing/mapped/test",
+                "preprocessing/mapped/test/test.sorted.cram",
+                "preprocessing/mapped/test/test.sorted.cram.crai",
+                "reference",
+                "reference/bwa",
+                "reference/bwa/genome.amb",
+                "reference/bwa/genome.ann",
+                "reference/bwa/genome.bwt",
+                "reference/bwa/genome.pac",
+                "reference/bwa/genome.sa",
+                "reference/intervals",
+                "reference/intervals/chr22_1-40001.bed",
+                "reference/intervals/chr22_1-40001.bed.gz",
+                "reference/intervals/genome.bed",
+                "reference/intervals/genome.bed.gz"
+            ],
+            [
+                "genome.amb:md5,1891c1de381b3a96d4e72f590fde20c1",
+                "genome.ann:md5,2df4aa2d7580639fa0fcdbcad5e2e969",
+                "genome.bwt:md5,815eded87e4cb6b0f1daab5c4d6e30af",
+                "genome.pac:md5,8569fbdb2c98c6fb16dfa73d8eacb070",
+                "genome.sa:md5,e7cff62b919448a3a3d0fe4aaf427594",
+                "chr22_1-40001.bed:md5,87a15eb9c2ff20ccd5cd8735a28708f7",
+                "chr22_1-40001.bed.gz:md5,87a15eb9c2ff20ccd5cd8735a28708f7",
+                "genome.bed:md5,a87dc7d20ebca626f65cc16ff6c97a3e",
+                "genome.bed.gz:md5,a87dc7d20ebca626f65cc16ff6c97a3e"
+            ],
+            [
+                [
+                    "test.sorted.cram",
+                    "5534c350547fd253f0f2b9450362bed"
+                ]
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.09.0"
+        },
+        "timestamp": "2024-10-08T11:22:13.394114"
+    },
+    "Run with profile test | --aligner bwa-mem --save_reference --build_only_index": {
+        "content": [
+            5,
+            {
+                "Workflow": {
+                    "nf-core/sarek": "v3.5.0dev"
+                }
+            },
+            [
+                "csv",
+                "pipeline_info",
+                "pipeline_info/nf_core_sarek_software_mqc_versions.yml",
+                "reference",
+                "reference/bwa",
+                "reference/bwa/genome.amb",
+                "reference/bwa/genome.ann",
+                "reference/bwa/genome.bwt",
+                "reference/bwa/genome.pac",
+                "reference/bwa/genome.sa",
+                "reference/intervals",
+                "reference/intervals/chr22_1-40001.bed",
+                "reference/intervals/chr22_1-40001.bed.gz",
+                "reference/intervals/genome.bed",
+                "reference/intervals/genome.bed.gz"
+            ],
+            [
+                "genome.amb:md5,1891c1de381b3a96d4e72f590fde20c1",
+                "genome.ann:md5,2df4aa2d7580639fa0fcdbcad5e2e969",
+                "genome.bwt:md5,815eded87e4cb6b0f1daab5c4d6e30af",
+                "genome.pac:md5,8569fbdb2c98c6fb16dfa73d8eacb070",
+                "genome.sa:md5,e7cff62b919448a3a3d0fe4aaf427594",
+                "chr22_1-40001.bed:md5,87a15eb9c2ff20ccd5cd8735a28708f7",
+                "chr22_1-40001.bed.gz:md5,87a15eb9c2ff20ccd5cd8735a28708f7",
+                "genome.bed:md5,a87dc7d20ebca626f65cc16ff6c97a3e",
+                "genome.bed.gz:md5,a87dc7d20ebca626f65cc16ff6c97a3e"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.9.0",
+            "nextflow": "24.09.0"
+        },
+        "timestamp": "2024-10-08T11:23:05.203586"
+    }
+}

tests/aligner-bwa-mem2.nf.test

@@ -0,0 +1,81 @@
+nextflow_pipeline {
+
+    name "Test pipeline"
+    script "../main.nf"
+    tag "pipeline"
+    tag "pipeline_sarek"
+
+    test("Run with profile test | --aligner bwa-mem2 --save_reference | skip QC/recal/md") {
+
+        when {
+            params {
+                aligner = 'bwa-mem2'
+                modules_testdata_base_path = 'https://raw.githubusercontent.com/nf-core/test-datasets/modules/data/'
+                outdir = "$outputDir"
+                save_reference = true
+                skip_tools = 'baserecalibrator,fastqc,markduplicates,mosdepth,multiqc,samtools'
+                tools = ''
+            }
+        }
+
+        then {
+            // stable_name: All files + folders in ${params.outdir}/ with a stable name
+            def stable_name = getAllFilesFromDir(params.outdir, relative: true, includeDir: true, ignore: ['pipeline_info/*.{html,json,txt}'])
+            // stable_path: All files in ${params.outdir}/ with stable content
+            def stable_path = getAllFilesFromDir(params.outdir, ignoreFile: 'tests/.nftignore')
+            // cram_files: All cram files
+            def cram_files  = getAllFilesFromDir(params.outdir, include: ['**/*.cram'])
+            def fasta       = params.modules_testdata_base_path + 'genomics/homo_sapiens/genome/genome.fasta'
+            assertAll(
+                { assert workflow.success},
+                { assert snapshot(
+                    // Number of successful tasks
+                    workflow.trace.succeeded().size(),
+                    // pipeline versions.yml file for multiqc from which Nextflow version is removed because we tests pipelines on multiple Nextflow versions
+                    removeNextflowVersion("$outputDir/pipeline_info/nf_core_sarek_software_mqc_versions.yml"),
+                    // All stable path name, with a relative path
+                    stable_name,
+                    // All files with stable contents
+                    stable_path,
+                    // All cram files
+                    cram_files.collect{ file -> [ file.getName(), cram(file.toString(), fasta).getReadsMD5() ] }
+                ).match() }
+            )
+        }
+    }
+
+    test("Run with profile test | --aligner bwa-mem2 --save_reference --build_only_index") {
+
+        when {
+            params {
+                aligner = 'bwa-mem2'
+                build_only_index = true
+                input = false
+                outdir = "$outputDir"
+                save_reference = true
+                skip_tools = 'multiqc'
+                tools = ''
+            }
+        }
+
+        then {
+            // stable_name: All files + folders in ${params.outdir}/ with a stable name
+            def stable_name = getAllFilesFromDir(params.outdir, relative: true, includeDir: true, ignore: ['pipeline_info/*.{html,json,txt}'])
+            // stable_path: All files in ${params.outdir}/ with stable content
+            def stable_path = getAllFilesFromDir(params.outdir, ignoreFile: 'tests/.nftignore')
+            assertAll(
+                { assert workflow.success},
+                { assert snapshot(
+                    // Number of successful tasks
+                    workflow.trace.succeeded().size(),
+                    // pipeline versions.yml file for multiqc from which Nextflow version is removed because we tests pipelines on multiple Nextflow versions
+                    removeNextflowVersion("$outputDir/pipeline_info/nf_core_sarek_software_mqc_versions.yml"),
+                    // All stable path name, with a relative path
+                    stable_name,
+                    // All files with stable contents
+                    stable_path
+                ).match() }
+            )
+        }
+    }
+}