Merge pull request #6 from cmungall/linkml-issue-1373

throw meaningful exception when values > columns.

Author: cmungall

src/json_flattener/flattener.py

@@ -21,6 +21,14 @@
 OBJECT = Dict[KEYNAME, Any]
 
 
+class MissingColumnError(ValueError):
+    """Exception raised when a column is missing from a row."""
+
+    def __init__(self, row: ROW, value: Any):
+        """Initialize exception."""
+        super().__init__(f"Value {value} has no column row {row}")
+
+
 @unique
 class Serializer(Enum):
     """Vocabulary of methods to use for serialization objects."""
@@ -441,6 +449,8 @@ def _getval(x: str) -> Optional[Any]:
     for row in r:
         nu_obj = {}
         for k, v in row.items():
+            if k is None:
+                raise MissingColumnError(row, v)
             key_config = config.key_configs.get(k, None)
             v = v.replace("\\n", "\n").replace("\\t", "\t")
             # lists are demarcated by list markers

tests/inputs/badly-formatted.tsv

@@ -0,0 +1,2 @@
+source_language	translation_language	subject_id	predicate_id	source_value	translation_value	translation_status
+en	tr	HP:0004432	IAO:0000115	ab	"a	b"	CANDIDATE

tests/test_flattener.py

@@ -4,6 +4,7 @@
 import json
 import logging
 import unittest
+from pathlib import Path
 from typing import Any, List
 
 import yaml
@@ -17,7 +18,8 @@
     unflatten,
     unflatten_from_csv,
 )
-from tests import INPUT
+from json_flattener.flattener import MissingColumnError
+from tests import INPUT, INPUT_DIR
 
 
 def _json(obj) -> str:
@@ -392,6 +394,14 @@ def test_roundtrip_from_file(self):
         logging.info(roundtrip_json)
         self._roundtrip_to_tsv(objs, config=config)
 
+    def test_badly_formatted(self):
+        """
+        Tests graceful failure on badly formatted TSV input.
+        """
+        with open(str(Path(INPUT_DIR) / "badly-formatted.tsv")) as stream:
+            with self.assertRaises(MissingColumnError) as _e:
+                _objs = unflatten_from_csv(stream)
+
 
 if __name__ == "__main__":
     unittest.main()