Merge branch 'main' into feat-geodataframe-dtypes

chalmerlowe · web-flow · commit baa5981fa11f · 2025-05-14T09:29:18.000-04:00
diff --git a/.github/.OwlBot.lock.yaml b/.github/.OwlBot.lock.yaml
@@ -13,5 +13,5 @@
 # limitations under the License.
 docker:
   image: gcr.io/cloud-devrel-public-resources/owlbot-python:latest
-  digest: sha256:a7aef70df5f13313ddc027409fc8f3151422ec2a57ac8730fce8fa75c060d5bb
-# created: 2025-04-10T17:00:10.042601326Z
+  digest: sha256:3b3a31be60853477bc39ed8d9bac162cac3ba083724cecaad54eb81d4e4dae9c
+# created: 2025-04-16T22:40:03.123475241Z
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -5,6 +5,25 @@
 [1]: https://pypi.org/project/google-cloud-bigquery/#history
 
 
+## [3.32.0](https://github.com/googleapis/python-bigquery/compare/v3.31.0...v3.32.0) (2025-05-12)
+
+
+### Features
+
+* Add dataset access policy version attribute ([#2169](https://github.com/googleapis/python-bigquery/issues/2169)) ([b7656b9](https://github.com/googleapis/python-bigquery/commit/b7656b97c1bd6c204d0508b1851d114719686655))
+* Add preview support for incremental results ([#2145](https://github.com/googleapis/python-bigquery/issues/2145)) ([22b80bb](https://github.com/googleapis/python-bigquery/commit/22b80bba9d0bed319fd3102e567906c9b458dd02))
+* Add WRITE_TRUNCATE_DATA enum ([#2166](https://github.com/googleapis/python-bigquery/issues/2166)) ([4692747](https://github.com/googleapis/python-bigquery/commit/46927479085f13fd326e3f2388f60dfdd37f7f69))
+* Adds condition class and assoc. unit tests ([#2159](https://github.com/googleapis/python-bigquery/issues/2159)) ([a69d6b7](https://github.com/googleapis/python-bigquery/commit/a69d6b796d2edb6ba453980c9553bc9b206c5a6e))
+* Support BigLakeConfiguration (managed Iceberg tables) ([#2162](https://github.com/googleapis/python-bigquery/issues/2162)) ([a1c8e9a](https://github.com/googleapis/python-bigquery/commit/a1c8e9aaf60986924868d54a0ab0334e77002a39))
+* Update the AccessEntry class with a new condition attribute and unit tests ([#2163](https://github.com/googleapis/python-bigquery/issues/2163)) ([7301667](https://github.com/googleapis/python-bigquery/commit/7301667272dfbdd04b1a831418a9ad2d037171fb))
+
+
+### Bug Fixes
+
+* `query()` now warns when `job_id` is set and the default `job_retry` is ignored ([#2167](https://github.com/googleapis/python-bigquery/issues/2167)) ([ca1798a](https://github.com/googleapis/python-bigquery/commit/ca1798aaee2d5905fe688d3097f8ee5c989da333))
+* Empty record dtypes ([#2147](https://github.com/googleapis/python-bigquery/issues/2147)) ([77d7173](https://github.com/googleapis/python-bigquery/commit/77d71736fcc006d3ab8f8ba17955ad5f06e21876))
+* Table iterator should not use bqstorage when page_size is not None ([#2154](https://github.com/googleapis/python-bigquery/issues/2154)) ([e89a707](https://github.com/googleapis/python-bigquery/commit/e89a707b162182ededbf94cc9a0f7594bc2be475))
+
 ## [3.31.0](https://github.com/googleapis/python-bigquery/compare/v3.30.0...v3.31.0) (2025-03-20)
 
 
diff --git a/google/cloud/bigquery/client.py b/google/cloud/bigquery/client.py
@@ -1389,6 +1389,7 @@ def update_table(
         self,
         table: Table,
         fields: Sequence[str],
+        autodetect_schema: bool = False,
         retry: retries.Retry = DEFAULT_RETRY,
         timeout: TimeoutType = DEFAULT_TIMEOUT,
     ) -> Table:
@@ -1419,6 +1420,10 @@ def update_table(
             fields (Sequence[str]):
                 The fields of ``table`` to change, spelled as the
                 :class:`~google.cloud.bigquery.table.Table` properties.
+            autodetect_schema (bool):
+                Specifies if the schema of the table should be autodetected when
+                updating the table from the underlying source. Only applicable
+                for external tables.
             retry (Optional[google.api_core.retry.Retry]):
                 A description of how to retry the API call.
             timeout (Optional[float]):
@@ -1438,12 +1443,18 @@ def update_table(
         path = table.path
         span_attributes = {"path": path, "fields": fields}
 
+        if autodetect_schema:
+            query_params = {"autodetect_schema": True}
+        else:
+            query_params = {}
+
         api_response = self._call_api(
             retry,
             span_name="BigQuery.updateTable",
             span_attributes=span_attributes,
             method="PATCH",
             path=path,
+            query_params=query_params,
             data=partial,
             headers=headers,
             timeout=timeout,
diff --git a/google/cloud/bigquery/dataset.py b/google/cloud/bigquery/dataset.py
@@ -512,7 +512,7 @@ def __ne__(self, other):
         return not self == other
 
     def __repr__(self):
-        return f"<AccessEntry: role={self.role}, {self._entity_type}={self.entity_id}>"
+        return f"<AccessEntry: role={self.role}, {self.entity_type}={self.entity_id}>"
 
     def _key(self):
         """A tuple key that uniquely describes this field.
@@ -531,7 +531,7 @@ def _key(self):
             properties["condition"] = condition_key
 
         prop_tup = tuple(sorted(properties.items()))
-        return (self.role, self._entity_type, self.entity_id, prop_tup)
+        return (self.role, self.entity_type, self.entity_id, prop_tup)
 
     def __hash__(self):
         return hash(self._key())
diff --git a/google/cloud/bigquery/version.py b/google/cloud/bigquery/version.py
@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__version__ = "3.31.0"
+__version__ = "3.32.0"
diff --git a/owlbot.py b/owlbot.py
@@ -109,16 +109,6 @@
 
 python.py_samples()
 
-s.replace(
-    "docs/conf.py",
-    r'\{"members": True\}',
-    '{"members": True, "inherited-members": True}',
-)
-s.replace(
-    "docs/conf.py",
-    r"exclude_patterns = \[",
-    '\\g<0>\n    "google/cloud/bigquery_v2/**",  # Legacy proto-based types.',
-)
 s.replace(
     "samples/**/noxfile.py",
     'BLACK_VERSION = "black==22.3.0"',
diff --git a/samples/desktopapp/requirements-test.txt b/samples/desktopapp/requirements-test.txt
@@ -1,4 +1,4 @@
-google-cloud-testutils==1.6.0
+google-cloud-testutils==1.6.2
 pytest==8.3.5
 mock==5.2.0
 pytest-xdist==3.6.1
diff --git a/samples/desktopapp/requirements.txt b/samples/desktopapp/requirements.txt
@@ -1,2 +1,2 @@
 google-cloud-bigquery==3.31.0
-google-auth-oauthlib==1.2.1
+google-auth-oauthlib==1.2.2
diff --git a/samples/geography/requirements.txt b/samples/geography/requirements.txt
@@ -1,30 +1,31 @@
 attrs==25.3.0
-certifi==2025.1.31
+certifi==2025.4.26
 cffi==1.17.1
-charset-normalizer==3.4.1
-click==8.1.8
+charset-normalizer==3.4.2
+click===8.1.8; python_version == '3.9'
+click==8.2.0; python_version >= '3.10'
 click-plugins==1.1.1
 cligj==0.7.2
-db-dtypes==1.4.2
+db-dtypes==1.4.3
 Fiona==1.10.1
 geojson==3.2.0
 geopandas==1.0.1
 google-api-core==2.24.2
-google-auth==2.38.0
+google-auth==2.40.1
 google-cloud-bigquery==3.31.0
-google-cloud-bigquery-storage==2.30.0
+google-cloud-bigquery-storage==2.31.0
 google-cloud-core==2.4.3
 google-crc32c==1.7.1
 google-resumable-media==2.7.2
-googleapis-common-protos==1.69.2
+googleapis-common-protos==1.70.0
 grpcio==1.71.0
 idna==3.10
 munch==4.0.0
-mypy-extensions==1.0.0
-packaging==24.2
+mypy-extensions==1.1.0
+packaging==25.0
 pandas==2.2.3
 proto-plus==1.26.1
-pyarrow==19.0.1
+pyarrow==20.0.0
 pyasn1==0.6.1
 pyasn1-modules==0.4.2
 pycparser==2.22
@@ -33,10 +34,10 @@ python-dateutil==2.9.0.post0
 pytz==2025.2
 PyYAML==6.0.2
 requests==2.32.3
-rsa==4.9
+rsa==4.9.1
 Shapely===2.0.7; python_version == '3.9'
 Shapely==2.1.0; python_version >= '3.10'
 six==1.17.0
-typing-extensions==4.13.1
+typing-extensions==4.13.2
 typing-inspect==0.9.0
-urllib3==2.3.0
+urllib3==2.4.0
diff --git a/samples/magics/requirements-test.txt b/samples/magics/requirements-test.txt
@@ -1,4 +1,4 @@
-google-cloud-testutils==1.6.0
+google-cloud-testutils==1.6.2
 pytest==8.3.5
 mock==5.2.0
 pytest-xdist==3.6.1
diff --git a/samples/magics/requirements.txt b/samples/magics/requirements.txt
@@ -1,6 +1,6 @@
 bigquery_magics==0.9.0
-db-dtypes==1.4.2
+db-dtypes==1.4.3
 google.cloud.bigquery==3.31.0
-google-cloud-bigquery-storage==2.30.0
+google-cloud-bigquery-storage==2.31.0
 ipython===8.18.1
 pandas==2.2.3
diff --git a/samples/notebooks/requirements-test.txt b/samples/notebooks/requirements-test.txt
@@ -1,4 +1,4 @@
-google-cloud-testutils==1.6.0
+google-cloud-testutils==1.6.2
 pytest==8.3.5
 mock==5.2.0
 pytest-xdist==3.6.1
diff --git a/samples/notebooks/requirements.txt b/samples/notebooks/requirements.txt
@@ -1,9 +1,9 @@
 bigquery-magics==0.9.0
-db-dtypes==1.4.2
+db-dtypes==1.4.3
 google-cloud-bigquery==3.31.0
-google-cloud-bigquery-storage==2.30.0
+google-cloud-bigquery-storage==2.31.0
 ipython===8.18.1; python_version == '3.9'
-ipython==9.0.2; python_version >= '3.10'
+ipython==9.2.0; python_version >= '3.10'
 matplotlib===3.9.2; python_version == '3.9'
-matplotlib==3.10.1; python_version >= '3.10'
+matplotlib==3.10.3; python_version >= '3.10'
 pandas==2.2.3
diff --git a/samples/snippets/requirements-test.txt b/samples/snippets/requirements-test.txt
@@ -1,5 +1,5 @@
 # samples/snippets should be runnable with no "extras"
-google-cloud-testutils==1.6.0
+google-cloud-testutils==1.6.2
 pytest==8.3.5
 mock==5.2.0
 pytest-xdist==3.6.1
diff --git a/tests/system/test_client.py b/tests/system/test_client.py
@@ -978,6 +978,53 @@ def test_update_table_constraints(self):
         )
         self.assertIsNone(reference_table3.table_constraints, None)
 
+    def test_update_table_autodetect_schema(self):
+        dataset = self.temp_dataset(_make_dataset_id("bq_update_table_test"))
+
+        # Create an external table, restrict schema to one field
+        TABLE_NAME = "test_table"
+        set_schema = [bigquery.SchemaField("username", "STRING", mode="NULLABLE")]
+        table_arg = Table(dataset.table(TABLE_NAME))
+
+        # Create an external_config and include it in the table arguments
+        external_config = bigquery.ExternalConfig(bigquery.ExternalSourceFormat.AVRO)
+        external_config.source_uris = SOURCE_URIS_AVRO
+        external_config.reference_file_schema_uri = REFERENCE_FILE_SCHEMA_URI_AVRO
+        external_config.schema = set_schema
+        table_arg.external_data_configuration = external_config
+
+        self.assertFalse(_table_exists(table_arg))
+
+        table = helpers.retry_403(Config.CLIENT.create_table)(table_arg)
+        self.to_delete.insert(0, table)
+        self.assertTrue(_table_exists(table))
+
+        self.assertEqual(table.schema, set_schema)
+
+        # Update table with schema autodetection
+        updated_table_arg = Table(dataset.table(TABLE_NAME))
+
+        # Update the external_config and include it in the table arguments
+        updated_external_config = copy.deepcopy(external_config)
+        updated_external_config.autodetect = True
+        updated_external_config.schema = None
+        updated_table_arg.external_data_configuration = updated_external_config
+
+        # PATCH call with autodetect_schema=True to trigger schema inference
+        updated_table = Config.CLIENT.update_table(
+            updated_table_arg, ["external_data_configuration"], autodetect_schema=True
+        )
+
+        # The updated table should have a schema inferred from the reference
+        # file, which has all four fields.
+        expected_schema = [
+            bigquery.SchemaField("username", "STRING", mode="NULLABLE"),
+            bigquery.SchemaField("tweet", "STRING", mode="NULLABLE"),
+            bigquery.SchemaField("timestamp", "STRING", mode="NULLABLE"),
+            bigquery.SchemaField("likes", "INTEGER", mode="NULLABLE"),
+        ]
+        self.assertEqual(updated_table.schema, expected_schema)
+
     @staticmethod
     def _fetch_single_page(table, selected_fields=None):
         iterator = Config.CLIENT.list_rows(table, selected_fields=selected_fields)
diff --git a/tests/unit/test_client.py b/tests/unit/test_client.py
@@ -2385,7 +2385,7 @@ def test_update_table(self):
             "resourceTags": {"123456789012/key": "value"},
         }
         conn.api_request.assert_called_once_with(
-            method="PATCH", data=sent, path="/" + path, timeout=7.5
+            method="PATCH", data=sent, path="/" + path, timeout=7.5, query_params={}
         )
         self.assertEqual(updated_table.description, table.description)
         self.assertEqual(updated_table.friendly_name, table.friendly_name)
@@ -2439,6 +2439,7 @@ def test_update_table_w_custom_property(self):
             path="/%s" % path,
             data={"newAlphaProperty": "unreleased property"},
             timeout=DEFAULT_TIMEOUT,
+            query_params={},
         )
         self.assertEqual(
             updated_table._properties["newAlphaProperty"], "unreleased property"
@@ -2475,6 +2476,7 @@ def test_update_table_only_use_legacy_sql(self):
             path="/%s" % path,
             data={"view": {"useLegacySql": True}},
             timeout=DEFAULT_TIMEOUT,
+            query_params={},
         )
         self.assertEqual(updated_table.view_use_legacy_sql, table.view_use_legacy_sql)
 
@@ -2567,9 +2569,10 @@ def test_update_table_w_query(self):
                 "schema": schema_resource,
             },
             timeout=DEFAULT_TIMEOUT,
+            query_params={},
         )
 
-    def test_update_table_w_schema_None(self):
+    def test_update_table_w_schema_None_autodetect_schema(self):
         # Simulate deleting schema:  not sure if back-end will actually
         # allow this operation, but the spec says it is optional.
         path = "projects/%s/datasets/%s/tables/%s" % (
@@ -2611,7 +2614,9 @@ def test_update_table_w_schema_None(self):
         with mock.patch(
             "google.cloud.bigquery.opentelemetry_tracing._get_final_span_attributes"
         ) as final_attributes:
-            updated_table = client.update_table(table, ["schema"])
+            updated_table = client.update_table(
+                table, ["schema"], autodetect_schema=True
+            )
 
         final_attributes.assert_called_once_with(
             {"path": "/%s" % path, "fields": ["schema"]}, client, None
@@ -2623,6 +2628,7 @@ def test_update_table_w_schema_None(self):
         sent = {"schema": {"fields": None}}
         self.assertEqual(req[1]["data"], sent)
         self.assertEqual(req[1]["path"], "/%s" % path)
+        self.assertEqual(req[1]["query_params"], {"autodetect_schema": True})
         self.assertEqual(len(updated_table.schema), 0)
 
     def test_update_table_delete_property(self):
diff --git a/tests/unit/test_dataset.py b/tests/unit/test_dataset.py
@@ -613,6 +613,15 @@ def test_equality_and_hash_without_condition(self):
         assert hash(entry1) == hash(entry2)
         assert hash(entry1) != hash(entry3)  # Usually true
 
+    def test_equality_and_hash_from_api_repr(self):
+        """Compare equal entries where one was created via from_api_repr."""
+        entry1 = AccessEntry("OWNER", "specialGroup", "projectOwners")
+        entry2 = AccessEntry.from_api_repr(
+            {"role": "OWNER", "specialGroup": "projectOwners"}
+        )
+        assert entry1 == entry2
+        assert hash(entry1) == hash(entry2)
+
     def test_equality_and_hash_with_condition(self, condition_1, condition_2):
         cond1a = Condition(
             condition_1.expression, condition_1.title, condition_1.description
@@ -746,6 +755,13 @@ def test_dataset_property_with_condition(self, condition_1):
         assert "dataset" in entry._properties
         assert "condition" in entry._properties
 
+    def test_repr_from_api_repr(self):
+        """Check that repr() includes the correct entity_type when the object is initialized from a dictionary."""
+        api_repr = {"role": "OWNER", "userByEmail": "owner@example.com"}
+        entry = AccessEntry.from_api_repr(api_repr)
+        entry_str = repr(entry)
+        assert entry_str == "<AccessEntry: role=OWNER, userByEmail=owner@example.com>"
+
 
 class TestDatasetReference(unittest.TestCase):
     @staticmethod
@@ -1097,6 +1113,34 @@ def test_ctor_explicit(self):
         self.assertIsNone(dataset.location)
         self.assertEqual(dataset.is_case_insensitive, False)
 
+    def test_access_entries_getter_from_api_repr(self):
+        """Check that `in` works correctly when Dataset is made via from_api_repr()."""
+        from google.cloud.bigquery.dataset import AccessEntry
+
+        dataset = self._get_target_class().from_api_repr(
+            {
+                "datasetReference": {"projectId": "my-proj", "datasetId": "my_dset"},
+                "access": [
+                    {
+                        "role": "OWNER",
+                        "userByEmail": "uilma@example.com",
+                    },
+                    {
+                        "role": "READER",
+                        "groupByEmail": "rhubbles@example.com",
+                    },
+                ],
+            }
+        )
+        assert (
+            AccessEntry("OWNER", "userByEmail", "uilma@example.com")
+            in dataset.access_entries
+        )
+        assert (
+            AccessEntry("READER", "groupByEmail", "rhubbles@example.com")
+            in dataset.access_entries
+        )
+
     def test_access_entries_setter_non_list(self):
         dataset = self._make_one(self.DS_REF)
         with self.assertRaises(TypeError):

Original file line number	Diff line number	Diff line change
`@@ -1,2 +1,2 @@`
`1`	`1`	`google-cloud-bigquery==3.31.0`
`2`		`-google-auth-oauthlib==1.2.1`
	`2`	`+google-auth-oauthlib==1.2.2`