Merge pull request #47 from pepkit/dev

Release v0.1.9

docs/changelog.md

@@ -1,7 +1,11 @@
 # Changelog
 
 This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html) and [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) format.
-
+## [0.1.9] - 2022-09-12
+### Fixed
+- CSV filter bug
+### Added
+- New test cases
 ## [0.1.8] - 2022-08-29
 ### Changed
 - the way of merging tables for multiline output format from eido convert

eido/_version.py

@@ -1 +1 @@
-__version__ = "0.1.8"
+__version__ = "0.1.9"

eido/output_formatters.py

@@ -103,15 +103,16 @@ def _convert_sample_to_row(
             Representation of sample as a CSV row.
         """
         sample_row = []
-
         for attribute in sample_attributes:
 
-            if MultilineOutputFormatter._sample_attribute_is_list(sample, attribute):
+            if MultilineOutputFormatter._sample_attribute_is_list(
+                sample, attribute
+            ) and getattr(sample, attribute):
                 value = getattr(sample, attribute)[sample_index]
             else:
                 value = getattr(sample, attribute, "")
 
-            sample_row.append(value)
+            sample_row.append(value or "")
 
         return ",".join(sample_row)
 

tests/conftest.py

@@ -57,7 +57,7 @@ def schema_imports_file_path(schemas_path):
 
 @pytest.fixture
 def taxprofiler_project_path(peps_path):
-    return os.path.join(peps_path, "taxprofiler_pep", "config.yaml")
+    return os.path.join(peps_path, "multiline_output", "config.yaml")
 
 
 @pytest.fixture
@@ -67,7 +67,25 @@ def taxprofiler_project(taxprofiler_project_path):
 
 @pytest.fixture
 def path_to_taxprofiler_csv_multiline_output(peps_path):
-    return os.path.join(peps_path, "taxprofiler_pep", "multiline_output.csv")
+    return os.path.join(peps_path, "multiline_output", "multiline_output.csv")
+
+
+@pytest.fixture
+def path_pep_with_fasta_column(peps_path):
+    return os.path.join(peps_path, "pep_with_fasta_column", "config.yaml")
+
+
+@pytest.fixture
+def project_pep_with_fasta_column(path_pep_with_fasta_column):
+    return Project(path_pep_with_fasta_column, sample_table_index="sample")
+
+
+@pytest.fixture
+def output_pep_with_fasta_column(path_pep_with_fasta_column):
+    with open(
+        os.path.join(os.path.dirname(path_pep_with_fasta_column), "output.csv")
+    ) as f:
+        return f.read()
 
 
 @pytest.fixture
@@ -77,6 +95,24 @@ def taxprofiler_csv_multiline_output(path_to_taxprofiler_csv_multiline_output):
     )
 
 
+@pytest.fixture
+def path_pep_nextflow_taxprofiler(peps_path):
+    return os.path.join(peps_path, "pep_nextflow_taxprofiler", "config.yaml")
+
+
+@pytest.fixture
+def project_pep_nextflow_taxprofiler(path_pep_nextflow_taxprofiler):
+    return Project(path_pep_nextflow_taxprofiler, sample_table_index="sample")
+
+
+@pytest.fixture
+def output_pep_nextflow_taxprofiler(path_pep_nextflow_taxprofiler):
+    with open(
+        os.path.join(os.path.dirname(path_pep_nextflow_taxprofiler), "output.csv")
+    ) as f:
+        return f.read()
+
+
 @pytest.fixture
 def save_result_mock(mocker):
     return mocker.patch("eido.conversion.save_result")

tests/data/peps/pep_nextflow_taxprofiler/config.yaml

@@ -0,0 +1,3 @@
+pep_version: "2.1.0"
+sample_table: "samplesheet.csv"
+

tests/data/peps/pep_nextflow_taxprofiler/output.csv

@@ -0,0 +1,7 @@
+sample,instrument_platform,run_accession,fastq_1,fastq_2,fasta
+2611,ILLUMINA,ERR5766174,,,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fasta/ERX5474930_ERR5766174_1.fa.gz
+2613,ILLUMINA,ERR5766181,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474937_ERR5766181_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474937_ERR5766181_2.fastq.gz,
+ERR3201952,OXFORD_NANOPORE,ERR3201952,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERR3201952.fastq.gz,,
+2612,ILLUMINA,ERR5766176,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_2.fastq.gz,
+2612,ILLUMINA,ERR5766176_B,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_B_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_B_2.fastq.gz,
+2612,ILLUMINA,ERR5766180,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474936_ERR5766180_1.fastq.gz,,

tests/data/peps/pep_nextflow_taxprofiler/samplesheet.csv

@@ -0,0 +1,7 @@
+sample,instrument_platform,run_accession,fastq_1,fastq_2,fasta
+2611,ILLUMINA,ERR5766174,,,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fasta/ERX5474930_ERR5766174_1.fa.gz
+2612,ILLUMINA,ERR5766176,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_2.fastq.gz,
+2612,ILLUMINA,ERR5766176_B,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_B_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_B_2.fastq.gz,
+2612,ILLUMINA,ERR5766180,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474936_ERR5766180_1.fastq.gz,,
+2613,ILLUMINA,ERR5766181,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474937_ERR5766181_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474937_ERR5766181_2.fastq.gz,
+ERR3201952,OXFORD_NANOPORE,ERR3201952,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERR3201952.fastq.gz,,

tests/data/peps/pep_with_fasta_column/config.yaml

@@ -0,0 +1,3 @@
+pep_version: "2.0.0"
+sample_table: "samplesheet.csv"
+subsample_table: "subsamplesheet.csv"

tests/data/peps/pep_with_fasta_column/output.csv

@@ -0,0 +1,8 @@
+sample,strandedness,instrument_platform,run_accession,fastq_1,fastq_2,fasta
+WT_REP1,reverse,ABI_SOLID,runaccession1,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357076_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357076_2.fastq.gz,
+WT_REP1,reverse,ABI_SOLID,runaccession2,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357071_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357071_2.fastq.gz,
+WT_REP2,reverse,BGISEQ,123123123,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357072_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357072_2.fastq.gz,
+RAP1_UNINDUCED_REP1,reverse,CAPILLARY,somerunaccesion,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357073_1.fastq.gz,,
+RAP1_UNINDUCED_REP2,reverse,COMPLETE_GENOMICS,ERR2412421,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357074_1.fastq.gz,,
+RAP1_UNINDUCED_REP2,reverse,COMPLETE_GENOMICS,xxxxxxxxxx,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357075_1.fastq.gz,,
+RAP1_IAA_30M_REP1,reverse,DNBSEQ,None,,,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fasta/ERX5474930_ERR5766174_1.fa.gz

tests/data/peps/pep_with_fasta_column/samplesheet.csv

@@ -0,0 +1,6 @@
+sample,strandedness,instrument_platform
+WT_REP1,reverse,ABI_SOLID
+WT_REP2,reverse,BGISEQ
+RAP1_UNINDUCED_REP1,reverse,CAPILLARY
+RAP1_UNINDUCED_REP2,reverse,COMPLETE_GENOMICS
+RAP1_IAA_30M_REP1,reverse,DNBSEQ

tests/data/peps/pep_with_fasta_column/subsamplesheet.csv

@@ -0,0 +1,8 @@
+sample,run_accession,fastq_1,fastq_2,fasta,strandedness
+WT_REP1,runaccession1,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357076_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357076_2.fastq.gz,,reverse
+WT_REP1,runaccession2,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357071_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357071_2.fastq.gz,,reverse
+WT_REP2,123123123,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357072_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357072_2.fastq.gz,,reverse
+RAP1_UNINDUCED_REP1,somerunaccesion,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357073_1.fastq.gz,,,reverse
+RAP1_UNINDUCED_REP2,ERR2412421,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357074_1.fastq.gz,,,reverse
+RAP1_UNINDUCED_REP2,xxxxxxxxxx,https://raw.githubusercontent.com/nf-core/test-datasets/rnaseq/testdata/GSE110004/SRR6357075_1.fastq.gz,,,reverse
+RAP1_IAA_30M_REP1,None,,,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fasta/ERX5474930_ERR5766174_1.fa.gz,reverse

tests/test_conversions.py

@@ -33,7 +33,6 @@ def test_basic_filter(self, save_result_mock, project_object):
     def test_csv_filter(
         self, save_result_mock, taxprofiler_project, taxprofiler_csv_multiline_output
     ):
-
         conv_result = run_filter(
             taxprofiler_project,
             "csv",
@@ -43,3 +42,35 @@ def test_csv_filter(
 
         assert save_result_mock.called
         assert conv_result["samples"] == taxprofiler_csv_multiline_output
+
+    def test_csv_filter_handles_empty_fasta_correclty(
+        self,
+        project_pep_with_fasta_column,
+        output_pep_with_fasta_column,
+        save_result_mock,
+    ):
+        conv_result = run_filter(
+            project_pep_with_fasta_column,
+            "csv",
+            verbose=False,
+            plugin_kwargs={"paths": {"samples": "out/basic_prj.txt"}},
+        )
+
+        assert save_result_mock.called
+        assert conv_result == {"samples": output_pep_with_fasta_column}
+
+    def test_eido_csv_filter_filters_nextflow_taxprofiler_input_correctly(
+        self,
+        project_pep_nextflow_taxprofiler,
+        output_pep_nextflow_taxprofiler,
+        save_result_mock,
+    ):
+        conv_result = run_filter(
+            project_pep_nextflow_taxprofiler,
+            "csv",
+            verbose=False,
+            plugin_kwargs={"paths": {"samples": "out/basic_prj.txt"}},
+        )
+
+        assert save_result_mock.called
+        assert conv_result == {"samples": output_pep_nextflow_taxprofiler}