Merge pull request #567 from gouttegd/strip-trailing-tabs-in-header

Strip trailing tabs in embedded YAML header.

Author: matentzn

src/sssom/parsers.py

@@ -121,7 +121,11 @@ def _separate_metadata_and_table_from_stream(s: io.StringIO):
             if header_section:
                 header_section = False
         elif header_section:
-            metadata_component.write(line)
+            # We strip any trailing tabs. Such tabs may have been left
+            # by a spreadsheet editor who treated the header lines as
+            # if they were normal data lines; they would prevent the
+            # YAML parser from correctly parsing the metadata block.
+            metadata_component.write(line.rstrip("\t\n") + "\n")
         else:
             logging.info(
                 f"Line {line} is starting with hash symbol, but header section is already passed. "

tests/data/trailing-tabs.sssom.tsv

@@ -0,0 +1,21 @@
+#curie_map:					
+#  COMENT: https://example.com/entities/					
+#  COMPID: https://example.com/people/					
+#  ORGENT: https://example.org/entities/					
+#  ORGPID: https://example.org/people/					
+#mapping_set_id: https://example.org/sets/exo2c					
+#mapping_set_title: O2C set					
+#creator_id:					
+#  - ORGPID:0000-0000-0001-1234					
+#  - COMPID:0000-0000-0002-5678					
+#license: https://creativecommons.org/licenses/by/4.0/					
+#publication_date: 2023-09-13					
+subject_id	subject_label	predicate_id	object_id	object_label	mapping_justification
+ORGENT:0001	alice	skos:closeMatch	COMENT:0011	alpha	semapv:ManualMappingCuration
+ORGENT:0002	bob	skos:closeMatch	COMENT:0012	beta	semapv:ManualMappingCuration
+ORGENT:0004	daphne	skos:closeMatch	COMENT:0014	delta	semapv:ManualMappingCuration
+ORGENT:0005	eve	skos:closeMatch	COMENT:0015	epsilon	semapv:ManualMappingCuration
+ORGENT:0006	fanny	skos:closeMatch	COMENT:0016	zeta	semapv:ManualMappingCuration
+ORGENT:0007	gavin	skos:exactMatch	COMENT:0013	gamma	semapv:ManualMappingCuration
+ORGENT:0008	hector	skos:closeMatch	COMENT:0017	eta	semapv:ManualMappingCuration
+ORGENT:0009	ivan	skos:exactMatch	COMENT:0019	iota	semapv:ManualMappingCuration

tests/test_parsers.py

@@ -339,6 +339,17 @@ def test_parse_obographs_merged(self):
         msdf = parse_sssom_table(outfile)
         self.assertTrue(custom_curie_map.items() <= msdf.prefix_map.items())
 
+    def test_parse_trailing_tabs_in_metadata_header(self):
+        """Test parsing a file containing trailing tabs in header."""
+        input_path = f"{test_data_dir}/trailing-tabs.sssom.tsv"
+        msdf = parse_sssom_table(input_path)
+        self.assertEqual(msdf.metadata["mapping_set_id"], "https://example.org/sets/exo2c")
+        self.assertEqual(
+            len(msdf.df),
+            8,
+            f"{input_path} has the wrong number of mappings.",
+        )
+
 
 class TestParseExplicit(unittest.TestCase):
     """This test case contains explicit tests for parsing."""