Add sativaepang/* modules: reference, lootasks, looplace, looscore

modules/nf-core/sativaepang/looplace/environment.yml

@@ -0,0 +1,7 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - bioconda::sativa-epang=0.9.3.4

modules/nf-core/sativaepang/looplace/main.nf

@@ -0,0 +1,69 @@
+process SATIVAEPANG_LOOPLACE {
+    tag "$meta.id"
+    label 'process_high'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine in ['singularity', 'apptainer'] && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/sativa-epang:0.9.3.4--py314hab16a5f_0' :
+        'quay.io/biocontainers/sativa-epang:0.9.3.4--py314hab16a5f_0' }"
+
+    input:
+    tuple val(meta), path(taskdir)
+
+    output:
+    tuple val(meta), path(taskdir), emit: taskdir
+    tuple val("${task.process}"), val('sativaepang'), eval("sed -n 's#.*share/sativa-epang-\\([0-9.]*\\)-.*#\\1#p' \$(command -v sativa-epang)"), topic: versions, emit: versions_sativaepang
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    """
+    # \$taskdir is staged from sativaepang/lootasks's own output -- a symlink under
+    # local/shared-filesystem staging, but a real copy under stageInMode 'copy' (the
+    # default for cloud storage without Fusion). loo-place writes a jplace (and logs)
+    # inside every fold directory, which would mutate that other task's output in place
+    # and make this task uncacheable across -resume (same class of bug as
+    # nf-core/modules#12799). Give it a private, writable directory tree instead: fold
+    # subdirectories are real (new) directories here, their read-only contents
+    # (ref.nwk/ref.fasta/query.fasta) stay symlinks to avoid copying large alignments,
+    # and manifest.json at the top level is a plain symlink since loo-place never writes
+    # there. mv handles both staging modes: it moves a symlink as a symlink, and renames
+    # a real directory in place, so nothing is deleted either way.
+    mv "$taskdir" "${taskdir}.staged"
+    real_taskdir=\$(readlink -f "${taskdir}.staged")
+    mkdir "$taskdir"
+    find "\$real_taskdir" -mindepth 1 -maxdepth 1 | while read -r entry; do
+        name=\$(basename "\$entry")
+        if [ -d "\$entry" ]; then
+            mkdir "$taskdir/\$name"
+            find "\$entry" -mindepth 1 -maxdepth 1 -exec ln -s {} "$taskdir/\$name"/ \\;
+        else
+            ln -s "\$entry" "$taskdir/\$name"
+        fi
+    done
+
+    sativa-epang \\
+        -stage loo-place \\
+        -taskdir $taskdir \\
+        -T ${task.cpus} \\
+        ${args}
+    """
+
+    stub:
+    """
+    mv "$taskdir" "${taskdir}.staged"
+    real_taskdir=\$(readlink -f "${taskdir}.staged")
+    mkdir "$taskdir"
+    find "\$real_taskdir" -mindepth 1 -maxdepth 1 | while read -r entry; do
+        name=\$(basename "\$entry")
+        if [ -d "\$entry" ]; then
+            mkdir "$taskdir/\$name"
+            touch "$taskdir/\$name/epa_result.jplace"
+        else
+            ln -s "\$entry" "$taskdir/\$name"
+        fi
+    done
+    """
+}

modules/nf-core/sativaepang/looplace/meta.yml

@@ -0,0 +1,68 @@
+name: "sativaepang_looplace"
+description: Place every leave-one-out fold with EPA-ng via -stage loo-place, writing a jplace into each fold directory
+keywords:
+  - phylogenetics
+  - taxonomy
+  - mislabelling
+  - placement
+  - leave-one-out
+tools:
+  - "sativaepang":
+      description: "SATIVA with EPA-ng as the placement engine"
+      homepage: "https://github.com/Aaramis/sativa-epang"
+      documentation: "https://github.com/Aaramis/sativa-epang/blob/main/CHANGES-epa-ng.md"
+      tool_dev_url: "https://github.com/Aaramis/sativa-epang"
+      doi: "10.1093/nar/gkw396"
+      licence:
+        - "GPL-3.0-or-later"
+      identifier: ""
+input:
+  - - meta:
+        type: map
+        description: Groovy Map containing sample information. e.g. `[ id:'sample1' ]`
+    - taskdir:
+        type: directory
+        description: |
+          The `*.l1o_tasks` directory produced by `sativaepang/lootasks`: one
+          subdirectory per fold (`ref.nwk`, `ref.fasta`, `query.fasta`) plus a
+          `manifest.json`.
+        pattern: "*.l1o_tasks"
+output:
+  taskdir:
+    - - meta:
+          type: map
+          description: Groovy Map containing sample information. e.g. `[ id:'sample1' ]`
+      - taskdir:
+          type: directory
+          description: |
+            The same taskdir, with an `epa_result.jplace` (plus `epang.log` and
+            `epa_info.log`) added to every fold subdirectory. Input to
+            `sativaepang/looscore`.
+          pattern: "*.l1o_tasks"
+  versions_sativaepang:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - sativaepang:
+          type: string
+          description: The name of the tool
+      - sed -n 's#.*share/sativa-epang-\([0-9.]*\)-.*#\1#p' \$(command -v sativa-epang):
+          type: eval
+          description: The expression to obtain the version of the tool
+
+topics:
+  versions:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - sativaepang:
+          type: string
+          description: The name of the tool
+      - sed -n 's#.*share/sativa-epang-\([0-9.]*\)-.*#\1#p' \$(command -v sativa-epang):
+          type: eval
+          description: The expression to obtain the version of the tool
+
+authors:
+  - "@erikrikarddaniel"
+maintainers:
+  - "@erikrikarddaniel"

modules/nf-core/sativaepang/looplace/tests/main.nf.test

@@ -0,0 +1,134 @@
+nextflow_process {
+
+    name "Test Process SATIVAEPANG_LOOPLACE"
+    script "../main.nf"
+    process "SATIVAEPANG_LOOPLACE"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "sativaepang"
+    tag "sativaepang/lootasks"
+    tag "sativaepang/looplace"
+
+    test("sativa-epang tiny example") {
+
+        setup {
+            run("SATIVAEPANG_LOOTASKS") {
+                script "../../lootasks/main.nf"
+                process {
+                    """
+                    input[0] = [
+                        [ id:'test' ],
+                        file(params.modules_testdata_base_path + 'delete_me/sativaepang/sativaepang_tiny.refjson', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'delete_me/sativaepang/sativaepang_tiny.model', checkIfExists: true)
+                    ]
+                    """
+                }
+            }
+        }
+
+        when {
+            process {
+                """
+                input[0] = SATIVAEPANG_LOOTASKS.out.taskdir
+                """
+            }
+        }
+
+        then {
+            // Every fold in the returned taskdir must have a jplace, not just process.success.
+            // (Separately verified by hand, not expressible cleanly here: the upstream
+            // SATIVAEPANG_LOOTASKS task's own directory on disk is left untouched -- this
+            // module stages a private, writable copy rather than mutating the upstream
+            // task's output in place, the same class of -resume-breaking bug fixed in
+            // nf-core/modules#12799.)
+            // EPA-ng's placement likelihoods aren't byte-reproducible between the conda
+            // environment and the container builds (confirmed: identical logs/inputs,
+            // differing epa_result.jplace content). epa_info.log/epang.log also embed a
+            // wall-clock "Elapsed Time: Ns" line that flips between runs under load, on any
+            // engine. Mask all three in the snapshot -- same precedent as
+            // sigprofiler/snapaligner's unstablePatterns usage -- and rely on the structural
+            // asserts below for the guarantee this module actually makes.
+            def taskdir = file(process.out.taskdir[0][1])
+            def foldDirs = taskdir.listFiles().findAll { it.isDirectory() && it.name.startsWith("fold_") }
+            assertAll(
+                { assert process.success },
+                { assert foldDirs.size() > 0 },
+                { assert foldDirs.every { new File(it.toString(), "epa_result.jplace").exists() } },
+                { assert snapshot(sanitizeOutput(process.out, unstablePatterns: ["**/epa_result.jplace", "**/epa_info.log", "**/epang.log"])).match() }
+            )
+        }
+    }
+
+    test("gtdb archaea 16S") {
+
+        setup {
+            run("SATIVAEPANG_LOOTASKS") {
+                script "../../lootasks/main.nf"
+                process {
+                    """
+                    input[0] = [
+                        [ id:'test' ],
+                        file(params.modules_testdata_base_path + 'delete_me/sativaepang/gtdb_archaea_16s.refjson', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'delete_me/sativaepang/gtdb_archaea_16s.model', checkIfExists: true)
+                    ]
+                    """
+                }
+            }
+        }
+
+        when {
+            process {
+                """
+                input[0] = SATIVAEPANG_LOOTASKS.out.taskdir
+                """
+            }
+        }
+
+        then {
+            def taskdir = file(process.out.taskdir[0][1])
+            def foldDirs = taskdir.listFiles().findAll { it.isDirectory() && it.name.startsWith("fold_") }
+            assertAll(
+                { assert process.success },
+                { assert foldDirs.size() > 0 },
+                { assert foldDirs.every { new File(it.toString(), "epa_result.jplace").exists() } },
+                { assert snapshot(sanitizeOutput(process.out, unstablePatterns: ["**/epa_result.jplace", "**/epa_info.log", "**/epang.log"])).match() }
+            )
+        }
+    }
+
+    test("sativa-epang tiny example - stub") {
+
+        options "-stub"
+
+        setup {
+            run("SATIVAEPANG_LOOTASKS") {
+                script "../../lootasks/main.nf"
+                process {
+                    """
+                    input[0] = [
+                        [ id:'test' ],
+                        file(params.modules_testdata_base_path + 'delete_me/sativaepang/sativaepang_tiny.refjson', checkIfExists: true),
+                        file(params.modules_testdata_base_path + 'delete_me/sativaepang/sativaepang_tiny.model', checkIfExists: true)
+                    ]
+                    """
+                }
+            }
+        }
+
+        when {
+            process {
+                """
+                input[0] = SATIVAEPANG_LOOTASKS.out.taskdir
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+    }
+}