change behavior to use shorthand columns if extant even if other defaults defined

Author: AmandaBirmingham

metameq/src/metadata_extender.py

@@ -728,9 +728,10 @@ def _find_internal_col_source_name(
     if specified_name:
         if specified_name != internal_key:
             if internal_key in raw_metadata_df.columns:
-                raise ValueError(f"Metadata contains both '{internal_key}' and "
-                                 f"'{specified_name}' columns, which are both specified as "
-                                 f"names for the {internal_key} field.")
+                logger.warning(
+                    f"Metadata contains both '{internal_key}' and "
+                    f"'{specified_name}' columns; using '{internal_key}'.")
+                return None
             if specified_name not in raw_metadata_df.columns:
                 raise ValueError(
                     f"Specified {internal_key} column '{specified_name}' not found in metadata.")

metameq/tests/test_metadata_extender/test__extend_metadata_from_full_flat_config.py

@@ -258,19 +258,24 @@ def test_col_name_not_found_raises(self):
                 input_df, self.BASIC_FLAT_CONFIG, None,
                 "nonexistent_col", None)
 
-    def test_col_name_conflict_raises(self):
-        """Test that both internal and alternate columns raises ValueError."""
+    def test_col_name_conflict_warns_and_uses_internal_key(self):
+        """Test that both internal and alternate columns warns and uses internal key."""
         input_df = pandas.DataFrame({
             SAMPLE_NAME_KEY: ["sample1"],
             HOSTTYPE_SHORTHAND_KEY: ["human"],
             "host_type": ["human"],
             SAMPLETYPE_SHORTHAND_KEY: ["stool"]
         })
 
-        with self.assertRaisesRegex(ValueError, "contains both"):
-            _extend_metadata_from_full_flat_config(
-                input_df, self.BASIC_FLAT_CONFIG, None,
-                "host_type", None)
+        with self.assertLogs("metameq.src.metadata_extender", level="WARNING") as cm:
+            result_df, validation_msgs_df, col_name_mapping = \
+                _extend_metadata_from_full_flat_config(
+                    input_df, self.BASIC_FLAT_CONFIG, None,
+                    "host_type", None)
+
+        self.assertTrue(any("contains both" in msg for msg in cm.output))
+        self.assertEqual(HOSTTYPE_SHORTHAND_KEY,
+                         col_name_mapping[HOSTTYPE_SHORTHAND_KEY])
 
     def test_unknown_host_type(self):
         """Test that unknown host type adds QC note."""

metameq/tests/test_metadata_extender/test_group_column_identification.py

@@ -551,20 +551,23 @@ def test__find_internal_col_source_name_param_key_equals_internal_key(self):
 
         self.assertIsNone(result)
 
-    def test__find_internal_col_source_name_param_key_both_cols_raises(self):
-        """Test raises ValueError when both internal_key and param_key columns exist."""
+    def test__find_internal_col_source_name_param_key_both_cols_warns(self):
+        """Test warns and returns None when both internal_key and param_key columns exist."""
         input_df = pandas.DataFrame({
             "host_type": ["human"],
             HOSTTYPE_SHORTHAND_KEY: ["human"]
         })
         config_dict = {}
 
-        with self.assertRaisesRegex(ValueError, "contains both"):
-            _find_internal_col_source_name(
+        with self.assertLogs("metameq.src.metadata_extender", level="WARNING") as cm:
+            result = _find_internal_col_source_name(
                 input_df, config_dict,
                 "host_type", HOSTTYPE_SHORTHAND_KEY,
                 HOSTTYPE_COL_OPTIONS_KEY)
 
+        self.assertIsNone(result)
+        self.assertTrue(any("contains both" in msg for msg in cm.output))
+
     def test__find_internal_col_source_name_param_key_not_in_df_raises(self):
         """Test raises ValueError when param_key column not found in df."""
         input_df = pandas.DataFrame({
@@ -623,8 +626,8 @@ def test__find_internal_col_source_name_default_equals_internal_key(self):
 
         self.assertIsNone(result)
 
-    def test__find_internal_col_source_name_default_both_cols_raises(self):
-        """Test raises ValueError when default column and internal_key both exist in df."""
+    def test__find_internal_col_source_name_default_both_cols_warns(self):
+        """Test warns and returns None when default column and internal_key both exist in df."""
         input_df = pandas.DataFrame({
             "host_type": ["human"],
             HOSTTYPE_SHORTHAND_KEY: ["human"]
@@ -633,7 +636,10 @@ def test__find_internal_col_source_name_default_both_cols_raises(self):
             HOSTTYPE_COL_OPTIONS_KEY: ["host_type"]
         }
 
-        with self.assertRaisesRegex(ValueError, "contains both"):
-            _find_internal_col_source_name(
+        with self.assertLogs("metameq.src.metadata_extender", level="WARNING") as cm:
+            result = _find_internal_col_source_name(
                 input_df, config_dict,
                 None, HOSTTYPE_SHORTHAND_KEY, HOSTTYPE_COL_OPTIONS_KEY)
+
+        self.assertIsNone(result)
+        self.assertTrue(any("contains both" in msg for msg in cm.output))

metameq/tests/test_metadata_extender/test_group_populate_and_extend.py

@@ -1006,8 +1006,8 @@ def test_extend_metadata_df_col_name_not_found_raises(self):
                 input_df, study_config, None, None, self.TEST_STDS_FP,
                 hosttype_col_name="nonexistent_col")
 
-    def test_extend_metadata_df_col_name_conflicts_raises(self):
-        """Test that both internal and alternate columns existing raises ValueError."""
+    def test_extend_metadata_df_col_name_conflicts_warns_and_uses_internal_key(self):
+        """Test that both internal and alternate columns warns and uses internal key."""
         input_df = pandas.DataFrame({
             SAMPLE_NAME_KEY: ["sample1"],
             HOSTTYPE_SHORTHAND_KEY: ["human"],
@@ -1016,7 +1016,9 @@ def test_extend_metadata_df_col_name_conflicts_raises(self):
         })
         study_config = {}
 
-        with self.assertRaisesRegex(ValueError, "contains both"):
-            extend_metadata_df(
+        with self.assertLogs("metameq.src.metadata_extender", level="WARNING") as cm:
+            result_df, validation_msgs_df = extend_metadata_df(
                 input_df, study_config, None, None, self.TEST_STDS_FP,
                 hosttype_col_name="host_type")
+
+        self.assertTrue(any("contains both" in msg for msg in cm.output))