108 pairseq mismatch gz indextrue ignores key func

NEWS.rst

@@ -1,6 +1,10 @@
 Release Notes
 ================================================================================
 
+Version 0.7.8dev: Unreleased
+-------------------------------------------------------------------------------
+
+
 Version 0.7.7: January 19, 2026
 -------------------------------------------------------------------------------
 

presto/IO.py

@@ -124,7 +124,11 @@ def readSeqFile(seq_file, index=False, key_func=None):
                 # For gzip files, we can't use SeqIO.index directly, so we'll read into memory.
                 # SeqIO.index() cannot index regular .gz files, only BGZF format.
                 with openFile(seq_file, 'r') as handle:
-                    seq_records = {rec.id: rec for rec in SeqIO.parse(handle, seq_type)}
+                    # Use key_func if possible for consistent behavior with non-.gz indexing
+                    if key_func is None:
+                        seq_records = {rec.id: rec for rec in SeqIO.parse(handle, seq_type)}
+                    else:
+                        seq_records = {key_func(rec.id): rec for rec in SeqIO.parse(handle, seq_type)}
             else:
                 seq_records = SeqIO.index(seq_file, seq_type, key_function=key_func)
         else:

presto/Version.py

@@ -5,5 +5,5 @@
 __author__    = 'Jason Anthony Vander Heiden'
 __copyright__ = 'Copyright 2026 Kleinstein Lab, Yale University. All rights reserved.'
 __license__   = 'GNU Affero General Public License 3 (AGPL-3)'
-__version__   = '0.7.7'
-__date__      = '2026.01.19'
+__version__   = '0.7.8dev'
+__date__      = '2026.01.25'

tests/test_IO.py

@@ -140,6 +140,65 @@ def test_gzip_output_functionality(self):
             self.assertEqual(len(records), 1)
             print(f'Successfully read back record: {records[0].id}')
 
+    #@unittest.skip('-> test_readSeqFile_gzipped_with_key_func() skipped\n')
+    def test_readSeqFile_gzipped_with_key_func(self):
+        """
+        Test for issue #108: readSeqFile with .gz + index=True should respect key_func
+
+        When using .gz FASTQ with index=True, readSeqFile should use key_func 
+        to create keys instead of using rec.id directly. This is essential for 
+        PairSeq --coord presto to work with gzipped consensus files.
+
+        Issue: https://github.com/immcantation/presto/issues/108
+        """
+        # Test files with presto annotations (ID|CONSCOUNT=X|...)
+        presto_r1 = os.path.join(data_path, 'test_presto_R1.fastq.gz')
+        presto_r2 = os.path.join(data_path, 'test_presto_R2.fastq.gz')
+
+        if not os.path.exists(presto_r1) or not os.path.exists(presto_r2):
+            self.skipTest(f"Test files not found: {presto_r1} or {presto_r2}")
+
+        # Import getCoordKey to test with presto coordinate extraction
+        from presto.Annotation import getCoordKey
+
+        # Define key function that extracts presto ID (strips annotations after |)
+        def presto_key_func(header):
+            return getCoordKey(header, coord_type='presto')
+
+        # Read R1 file with index=True and key_func
+        r1_dict = readSeqFile(presto_r1, index=True, key_func=presto_key_func)
+        print(f'R1 sequences indexed: {len(r1_dict)}')
+        print(f'R1 keys: {list(r1_dict.keys())}')
+
+        # Read R2 file with index=True and key_func
+        r2_dict = readSeqFile(presto_r2, index=True, key_func=presto_key_func)
+        print(f'R2 sequences indexed: {len(r2_dict)}')
+        print(f'R2 keys: {list(r2_dict.keys())}')
+
+        # Verify that keys are clean IDs without annotations
+        # (not full headers like "SEQUENCE001|CONSCOUNT=2|...")
+        for key in r1_dict.keys():
+            self.assertNotIn('|', key, 
+                f"Key should not contain annotations: {key}")
+            self.assertNotIn('CONSCOUNT', key,
+                f"Key should not contain CONSCOUNT: {key}")
+
+        # Verify that both R1 and R2 have the same keys (pairable sequences)
+        self.assertEqual(set(r1_dict.keys()), set(r2_dict.keys()),
+            "R1 and R2 should have matching keys when using presto key_func")
+
+        # Expected keys based on our test data
+        expected_keys = {'SEQUENCE001', 'SEQUENCE002', 'SEQUENCE003'}
+        self.assertEqual(set(r1_dict.keys()), expected_keys,
+            f"Keys should be clean sequence IDs: {expected_keys}")
+
+        # Verify the original full headers are preserved in the records
+        for key, record in r1_dict.items():
+            self.assertIn('|', record.id,
+                f"Original record.id should contain full header with annotations")
+            self.assertTrue(record.id.startswith(key),
+                f"Record ID should start with the key: {record.id} vs {key}")
+
 
 if __name__ == '__main__':
     unittest.main()