ENH: adds transformer

q2_types/per_sample_sequences/_transformer.py

@@ -6,6 +6,7 @@
 # The full license is in the file LICENSE, distributed with this software.
 # ----------------------------------------------------------------------------
 
+import re
 import os
 import gzip
 import shutil
@@ -24,7 +25,8 @@
                CasavaOneEightSingleLanePerSampleDirFmt,
                CasavaOneEightLanelessPerSampleDirFmt,
                SingleEndFastqManifestPhred33, SingleEndFastqManifestPhred64,
-               PairedEndFastqManifestPhred33, PairedEndFastqManifestPhred64)
+               PairedEndFastqManifestPhred33, PairedEndFastqManifestPhred64,
+               QIIME1DemuxDirFmt)
 
 
 def _single_lane_per_sample_fastq_helper(dirfmt, output_cls, parse_lane=True):
@@ -342,3 +344,33 @@ def _9(fmt: PairedEndFastqManifestPhred64) \
     warnings.warn(_phred64_warning)
     return _fastq_manifest_helper(fmt, _write_phred64_to_phred33,
                                   single_end=False)
+
+
+@plugin.register_transformer
+def _12(dirfmt: SingleLanePerSampleSingleEndFastqDirFmt) \
+        -> QIIME1DemuxDirFmt:
+    with dirfmt.manifest.view(FastqManifestFormat).open() as fh:
+        input_manifest = _parse_and_validate_manifest(fh, single_end=True,
+                                                      absolute=False)
+
+    result = QIIME1DemuxDirFmt()
+    fp = str(result.path / 'seqs.fna')
+    with open(fp, 'w') as fh:
+        i = 0
+        for r in input_manifest.iterrows():
+            sample_id = r[1]['sample-id']
+            filename = r[1]['filename']
+            if re.search("\s", sample_id) is not None:
+                raise ValueError(
+                    "Whitespace was found in the ID for sample %s. Sample "
+                    "IDs with whitespace are incompatible with FASTA."
+                    % sample_id)
+            fq_reader = skbio.io.read('%s/%s' % (str(dirfmt), filename),
+                                      format='fastq', constructor=skbio.DNA,
+                                      phred_offset=33, verify=False)
+            for seq in fq_reader:
+                seq.metadata['id'] = '%s_%d' % (sample_id, i)
+                seq.write(fh)
+                i += 1
+
+    return result

q2_types/per_sample_sequences/tests/data/single-end-two-sample-data1/MANIFEST

@@ -0,0 +1,5 @@
+# Quantitative Insights Into Kneecaps and Armpits
+
+sample-id,filename,direction
+Human-Kneecap,Human-Kneecap_S1_L001_R1_001.fastq.gz,forward
+Human-Armpit,Human-Armpit_S2_L001_R1_001.fastq.gz,forward

q2_types/per_sample_sequences/tests/data/single-end-two-sample-data2/MANIFEST

@@ -0,0 +1,6 @@
+# Quantitative Insights Into Kneecaps and Armpits
+
+sample-id,filename,direction
+# space in sample-id
+Human-Kne ecap,Human-Kneecap_S1_L001_R1_001.fastq.gz,forward
+Human-Armpit,Human-Armpit_S2_L001_R1_001.fastq.gz,forward

q2_types/per_sample_sequences/tests/data/single-end-two-sample-data3/MANIFEST

@@ -0,0 +1,6 @@
+# Quantitative Insights Into Kneecaps and Armpits
+
+sample-id,filename,direction
+Human-Kneecap,Human-Kneecap_S1_L001_R1_001.fastq.gz,forward
+# tab in sample-id
+Human-Armp	it,Human-Armpit_S2_L001_R1_001.fastq.gz,forward

q2_types/per_sample_sequences/tests/test_transformer.py

@@ -23,7 +23,8 @@
     SingleEndFastqManifestPhred33,
     SingleEndFastqManifestPhred64,
     PairedEndFastqManifestPhred33,
-    PairedEndFastqManifestPhred64)
+    PairedEndFastqManifestPhred64,
+    QIIME1DemuxDirFmt)
 from q2_types.per_sample_sequences._transformer import (
     _validate_header,
     _validate_single_end_fastq_manifest_directions,
@@ -56,6 +57,62 @@ def test_slpspefdf_to_slpssefdf(self):
         for act, exp in zip(obs, expected):
             self.assertEqual(act, exp)
 
+    def test_slpssefdf_to_qiime1demuxdf(self):
+        filenames = ('single-end-two-sample-data1/MANIFEST',
+                     'metadata.yml',
+                     'Human-Kneecap_S1_L001_R1_001.fastq.gz',
+                     'Human-Armpit_S2_L001_R1_001.fastq.gz')
+        input, observed = self.transform_format(
+            SingleLanePerSampleSingleEndFastqDirFmt,
+            QIIME1DemuxDirFmt, filenames=filenames
+        )
+        expected1 = list(skbio.io.read(
+            '%s/Human-Kneecap_S1_L001_R1_001.fastq.gz' % str(input),
+            format='fastq', constructor=skbio.DNA
+        ))
+        expected2 = list(skbio.io.read(
+            '%s/Human-Armpit_S2_L001_R1_001.fastq.gz' % str(input),
+            format='fastq', constructor=skbio.DNA
+        ))
+        expected = \
+            list(zip(expected1, ['Human-Kneecap'] * len(expected1))) + \
+            list(zip(expected2, ['Human-Armpit'] * len(expected2)))
+        observed = skbio.io.read(
+            '%s/seqs.fna' % str(observed),
+            format='fasta', constructor=skbio.DNA
+        )
+        observed = list(observed)
+
+        self.assertEqual(len(observed), len(expected))
+
+        for i, obs in enumerate(observed):
+            # identifiers are as expected
+            self.assertEqual(obs.metadata['id'],
+                             '%s_%d' % (expected[i][1], i))
+            # sequences are as expected
+            self.assertEqual(str(obs), str(expected[i][0]))
+
+    def test_slpssefdf_to_qiime1demuxdf_bad_sample_ids(self):
+        filenames = ('single-end-two-sample-data2/MANIFEST',
+                     'metadata.yml',
+                     'Human-Kneecap_S1_L001_R1_001.fastq.gz',
+                     'Human-Armpit_S2_L001_R1_001.fastq.gz')
+        with self.assertRaisesRegex(ValueError,
+                                    expected_regex='space.*Human-K'):
+            self.transform_format(
+                SingleLanePerSampleSingleEndFastqDirFmt,
+                QIIME1DemuxDirFmt, filenames=filenames)
+
+        filenames = ('single-end-two-sample-data3/MANIFEST',
+                     'metadata.yml',
+                     'Human-Kneecap_S1_L001_R1_001.fastq.gz',
+                     'Human-Armpit_S2_L001_R1_001.fastq.gz')
+        with self.assertRaisesRegex(ValueError,
+                                    expected_regex='space.*Human-A'):
+            self.transform_format(
+                SingleLanePerSampleSingleEndFastqDirFmt,
+                QIIME1DemuxDirFmt, filenames=filenames)
+
     def test_casava_one_eight_single_lane_per_sample_dirfmt_to_slpssefdf(self):
         filenames = ('Human-Kneecap_S1_L001_R1_001.fastq.gz',)
         input, obs = self.transform_format(

setup.py

@@ -30,12 +30,16 @@
         'q2_types.feature_data.tests': ['data/*', 'data/taxonomy/*'],
         'q2_types.feature_table.tests': ['data/*'],
         'q2_types.ordination.tests': ['data/*'],
-        'q2_types.per_sample_sequences.tests': ['data/*',
-                                                'data/paired_end_data/*',
-                                                'data/single_end_data/*',
-                                                'data/absolute_manifests/*',
-                                                'data/relative_manifests/*',
-                                                'data/qiime1-demux-format/*'],
+        'q2_types.per_sample_sequences.tests':
+            ['data/*',
+             'data/paired_end_data/*',
+             'data/single_end_data/*',
+             'data/absolute_manifests/*',
+             'data/relative_manifests/*',
+             'data/qiime1-demux-format/*',
+             'data/single-end-two-sample-data1/*',
+             'data/single-end-two-sample-data2/*',
+             'data/single-end-two-sample-data3/*'],
         'q2_types.sample_data.tests': ['data/*'],
         'q2_types.tree.tests': ['data/*']
     },