table-validator 0.1.4__tar.gz → 0.1.6__tar.gz

{table_validator-0.1.4/table_validator.egg-info → table_validator-0.1.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: table-validator
-Version: 0.1.4
+Version: 0.1.6
 Summary: CLI tool for validating data migrations between Azure (Blob Storage / SQL Database) and Databricks Delta Lake catalogs
 License: MIT
 Requires-Python: >=3.9

{table_validator-0.1.4 → table_validator-0.1.6}/table_validator/validators/catalog_validator.py

@@ -398,6 +398,18 @@ class CatalogValidator:
             if pair_key in existing_pairs_lower:
                 continue
 
+            # An explicit mapping for this source schema supersedes any
+            # identical-name pair already seeded for it (e.g. a
+            # coincidental same-named schema on both sides) - the user
+            # named exactly one target for this source schema, so it must
+            # be validated once, against that target, not once against
+            # its own identical-name match AND again against the mapped
+            # target.
+            common_pairs = [
+                (s, t) for s, t in common_pairs if s.lower() != actual_source.lower()
+            ]
+            existing_pairs_lower = {(s.lower(), t.lower()) for s, t in common_pairs}
+
             common_pairs.append((actual_source, actual_target))
             existing_pairs_lower.add(pair_key)
             missing_set.discard(actual_source)
@@ -516,15 +528,18 @@ class CatalogValidator:
         names within an already-resolved source_schema/target_schema
         pair).
         """
-        common, missing, extra = self.compare_tables(
-            request.source_catalog, request.target_catalog, source_schema
-        )
-
-        # NOTE: compare_tables() above assumes the same schema name on
-        # both sides. When source_schema != target_schema (a schema_map
-        # pair), that identical-name baseline is meaningless - re-derive
-        # it directly from each side's own schema.
-        if source_schema.lower() != target_schema.lower():
+        if source_schema.lower() == target_schema.lower():
+            common, missing, extra = self.compare_tables(
+                request.source_catalog, request.target_catalog, source_schema
+            )
+        else:
+            # compare_tables() assumes the same schema name on both
+            # sides - calling it here would query the TARGET catalog for
+            # a schema named after the SOURCE schema (a schema_map pair
+            # means these differ), which doesn't exist there and raises
+            # SCHEMA_NOT_FOUND instead of just being a wasted query.
+            # Derive the baseline directly from each side's own,
+            # correctly-named schema instead.
             try:
                 source_tables = set(self.databricks.get_tables(request.source_catalog, source_schema))
             except Exception:
@@ -586,6 +601,18 @@ class CatalogValidator:
             if pair_key in existing_pairs_lower:
                 continue
 
+            # An explicit mapping for this source table supersedes any
+            # identical-name pair already seeded for it (e.g. a
+            # coincidental same-named table on both sides) - the user
+            # named exactly one target for this source table, so it must
+            # be validated once, against that target, not once against
+            # its own identical-name match AND again against the mapped
+            # target.
+            common_pairs = [
+                (s, t) for s, t in common_pairs if s.lower() != actual_source.lower()
+            ]
+            existing_pairs_lower = {(s.lower(), t.lower()) for s, t in common_pairs}
+
             common_pairs.append((actual_source, actual_target))
             existing_pairs_lower.add(pair_key)
             missing_set.discard(actual_source)

{table_validator-0.1.4 → table_validator-0.1.6/table_validator.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: table-validator
-Version: 0.1.4
+Version: 0.1.6
 Summary: CLI tool for validating data migrations between Azure (Blob Storage / SQL Database) and Databricks Delta Lake catalogs
 License: MIT
 Requires-Python: >=3.9

{table_validator-0.1.4 → table_validator-0.1.6}/table_validator.egg-info/scm_version.json

@@ -1,7 +1,7 @@
 {
-  "tag": "0.1.4",
+  "tag": "0.1.6",
   "distance": 0,
-  "node": "gb4d33e041e4eaea0e6645d64709a7fe7142c76db",
+  "node": "g9357213ed68fd98fc2c232aed2bb737f273cabbf",
   "dirty": false,
   "branch": "HEAD",
   "node_date": "2026-08-28"

{table_validator-0.1.4 → table_validator-0.1.6}/tests/test_catalog_validator.py

@@ -1473,6 +1473,36 @@ def test_table_map_validates_explicit_pair_with_different_names():
     connector.get_row_count.assert_called()
 
 
+def test_table_map_source_name_existing_identically_on_target_is_not_validated_twice():
+    """Real bug: the source table ('jd_example_data_2') coincidentally also
+    exists under that exact same name on the target side (an unrelated
+    leftover table), so compare_tables' plain intersection legitimately
+    finds an identical-name pair for it - in ADDITION to the table_map
+    entry pairing it with a different target ('jd_example_data_90'). The
+    mapping must supersede the identical-name match, not run alongside
+    it: one source table configured by the user must produce exactly one
+    validated table, against the mapped target only."""
+    connector = _make_connector()
+    connector.get_tables.side_effect = lambda catalog, schema: (
+        ["jd_example_data_2"] if catalog == "cat_source"
+        else ["jd_example_data_2", "jd_example_data_90"]
+    )
+    validator = CatalogValidator(connector)
+
+    result = validator.compare_catalogs(
+        _request(
+            tables=["jd_example_data_2"],
+            table_map={"jd_example_data_2": "jd_example_data_90"},
+        )
+    )
+
+    schema_result = result.schemas[0]
+    assert len(schema_result.tables) == 1
+    table = schema_result.tables[0]
+    assert table.table == "jd_example_data_90"
+    assert table.source_table_name == "jd_example_data_2"
+
+
 def test_table_map_to_nonexistent_target_table_produces_clear_error():
     """A table_map entry naming a target table that does NOT exist must
     produce a visible FAIL/ERROR with an informative message - never a
@@ -1544,6 +1574,48 @@ def test_schema_map_to_nonexistent_target_schema_produces_clear_error():
     assert result.status in (ValidationStatus.ERROR, ValidationStatus.FAIL)
 
 
+def test_schema_map_does_not_query_target_using_source_schema_name():
+    """Real bug: _resolve_table_pairs unconditionally called compare_tables
+    with the SOURCE schema name for both catalogs before checking whether
+    schema_map applied - querying the target catalog for a schema named
+    after the source schema, which doesn't exist there under a real
+    Databricks connector (raises SCHEMA_NOT_FOUND) rather than just being
+    a wasted no-op query. get_tables must only ever be called with each
+    catalog's own correctly-mapped schema name."""
+    connector = _make_connector()
+    connector.get_schemas.side_effect = lambda catalog: (
+        ["for_schema_validation"] if catalog == "cat_source" else ["bronze"]
+    )
+
+    def get_tables(catalog, schema):
+        if catalog == "cat_source" and schema == "for_schema_validation":
+            return ["bronze_fn_sku_add"]
+        if catalog == "cat_target" and schema == "bronze":
+            return ["bronze_fn_sku_add"]
+        raise RuntimeError(
+            f"Unable to list tables for '{catalog}.{schema}': SCHEMA_NOT_FOUND"
+        )
+
+    connector.get_tables.side_effect = get_tables
+    validator = CatalogValidator(connector)
+
+    result = validator.compare_catalogs(
+        _request(
+            schemas=["for_schema_validation"],
+            schema_map={"for_schema_validation": "bronze"},
+            tables=["bronze_fn_sku_add"],
+        )
+    )
+
+    assert result.status != ValidationStatus.ERROR
+    schema_result = result.schemas[0]
+    assert schema_result.schema_name == "bronze"
+    assert schema_result.status != ValidationStatus.ERROR
+    assert len(schema_result.tables) == 1
+    assert schema_result.tables[0].table == "bronze_fn_sku_add"
+    assert schema_result.tables[0].status in (ValidationStatus.PASS, ValidationStatus.FAIL)
+
+
 def test_no_map_configured_identical_name_discovery_unchanged():
     """Regression guard: with no schema_map/table_map set (the default,
     and today's only behavior), discovery must be completely unchanged -