deps: use pandas-gbq to determine schema in load_table_from_dataframe

.kokoro/docker/docs/requirements.txt

@@ -4,9 +4,9 @@
 #
 #    pip-compile --allow-unsafe --generate-hashes synthtool/gcp/templates/python_library/.kokoro/docker/docs/requirements.in
 #
-argcomplete==3.5.3 \
-    --hash=sha256:2ab2c4a215c59fd6caaff41a869480a23e8f6a5f910b266c1808037f4e375b61 \
-    --hash=sha256:c12bf50eded8aebb298c7b7da7a5ff3ee24dffd9f5281867dfe1424b58c55392
+argcomplete==3.5.2 \
+    --hash=sha256:036d020d79048a5d525bc63880d7a4b8d1668566b8a76daf1144c0bbe0f63472 \
+    --hash=sha256:23146ed7ac4403b70bd6026402468942ceba34a6732255b9edf5b7354f68a6bb
     # via nox
 colorlog==6.9.0 \
     --hash=sha256:5906e71acd67cb07a71e779c47c4bcb45fb8c2993eebe9e5adcd6a6f1b283eff \

google/cloud/bigquery/_pandas_helpers.py

@@ -12,7 +12,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-"""Shared helper functions for connecting BigQuery and pandas."""
+"""Shared helper functions for connecting BigQuery and pandas.
+
+NOTE: This module is DEPRECATED. Please make updates in the pandas-gbq package,
+instead. See: go/pandas-gbq-and-bigframes-redundancy and
+https://github.com/googleapis/python-bigquery-pandas/blob/main/pandas_gbq/schema/pandas_to_bigquery.py
+"""
 
 import concurrent.futures
 from datetime import datetime
@@ -39,6 +44,16 @@
 else:
     import numpy
 
+
+try:
+    import pandas_gbq.schema.pandas_to_bigquery  # type: ignore
+
+    pandas_gbq_import_exception = None
+except ImportError as exc:
+    pandas_gbq = None
+    pandas_gbq_import_exception = exc
+
+
 try:
     import db_dtypes  # type: ignore
 
@@ -429,6 +444,10 @@ def _first_array_valid(series):
 def dataframe_to_bq_schema(dataframe, bq_schema):
     """Convert a pandas DataFrame schema to a BigQuery schema.
 
+    DEPRECATED: Use
+    pandas_gbq.schema.pandas_to_bigquery.dataframe_to_bigquery_fields(),
+    instead. See: go/pandas-gbq-and-bigframes-redundancy.
+
     Args:
         dataframe (pandas.DataFrame):
             DataFrame for which the client determines the BigQuery schema.
@@ -444,6 +463,20 @@ def dataframe_to_bq_schema(dataframe, bq_schema):
             The automatically determined schema. Returns None if the type of
             any column cannot be determined.
     """
+    if pandas_gbq is None:
+        warnings.warn(
+            "Loading pandas DataFrame into BigQuery will require pandas-gbq "
+            "package version 0.26.0 or greater in future. "
+            f"Tried to import pandas-gbq and got: {pandas_gbq_import_exception}",
+            category=FutureWarning,
+        )
+    else:
+        return pandas_gbq.schema.pandas_to_bigquery.dataframe_to_bigquery_fields(
+            dataframe,
+            override_bigquery_fields=bq_schema,
+            index=True,
+        )
+
     if bq_schema:
         bq_schema = schema._to_schema_fields(bq_schema)
         bq_schema_index = {field.name: field for field in bq_schema}

google/cloud/bigquery/_pyarrow_helpers.py

@@ -12,7 +12,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-"""Shared helper functions for connecting BigQuery and pyarrow."""
+"""Shared helper functions for connecting BigQuery and pyarrow.
+
+NOTE: This module is DEPRECATED. Please make updates in the pandas-gbq package,
+instead. See: go/pandas-gbq-and-bigframes-redundancy and
+https://github.com/googleapis/python-bigquery-pandas/blob/main/pandas_gbq/schema/pyarrow_to_bigquery.py
+"""
 
 from typing import Any
 

pyproject.toml

@@ -74,6 +74,7 @@ bqstorage = [
 ]
 pandas = [
   "pandas >= 1.1.0",
+  "pandas-gbq >= 0.26.0",
   "pyarrow >= 3.0.0",
   "db-dtypes >= 0.3.0, < 2.0.0dev",
   "importlib_metadata >= 1.0.0; python_version < '3.8'",

testing/constraints-3.7.txt

@@ -26,6 +26,7 @@ opentelemetry-instrumentation==0.20b0
 opentelemetry-sdk==1.1.0
 packaging==20.0.0
 pandas==1.1.0
+pandas-gbq==0.26.0
 proto-plus==1.22.3
 protobuf==3.20.2
 pyarrow==3.0.0

tests/unit/test__pandas_helpers.py

@@ -34,6 +34,11 @@
 except ImportError:
     pandas = None
 
+try:
+    import pandas_gbq.schema.pandas_to_bigquery
+except ImportError:
+    pandas_gbq = None
+
 try:
     import geopandas
 except ImportError:
@@ -1280,7 +1285,21 @@ def test_dataframe_to_parquet_compression_method(module_under_test):
 
 
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
-def test_dataframe_to_bq_schema_w_named_index(module_under_test):
+@pytest.mark.skipif(pandas_gbq is None, reason="Requires `pandas-gbq`")
+def test_dataframe_to_bq_schema_returns_schema_with_pandas_gbq(
+    module_under_test, monkeypatch
+):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+    dataframe = pandas.DataFrame({"field00": ["foo", "bar"]})
+    got = module_under_test.dataframe_to_bq_schema(dataframe, [])
+    # Don't assert beyond this, since pandas-gbq is now source of truth.
+    assert got is not None
+
+
+@pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
+def test_dataframe_to_bq_schema_w_named_index(module_under_test, monkeypatch):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     df_data = collections.OrderedDict(
         [
             ("str_column", ["hello", "world"]),
@@ -1291,7 +1310,8 @@ def test_dataframe_to_bq_schema_w_named_index(module_under_test):
     index = pandas.Index(["a", "b"], name="str_index")
     dataframe = pandas.DataFrame(df_data, index=index)
 
-    returned_schema = module_under_test.dataframe_to_bq_schema(dataframe, [])
+    with pytest.warns(FutureWarning, match="pandas-gbq"):
+        returned_schema = module_under_test.dataframe_to_bq_schema(dataframe, [])
 
     expected_schema = (
         schema.SchemaField("str_index", "STRING", "NULLABLE"),
@@ -1303,7 +1323,9 @@ def test_dataframe_to_bq_schema_w_named_index(module_under_test):
 
 
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
-def test_dataframe_to_bq_schema_w_multiindex(module_under_test):
+def test_dataframe_to_bq_schema_w_multiindex(module_under_test, monkeypatch):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     df_data = collections.OrderedDict(
         [
             ("str_column", ["hello", "world"]),
@@ -1320,7 +1342,8 @@ def test_dataframe_to_bq_schema_w_multiindex(module_under_test):
     )
     dataframe = pandas.DataFrame(df_data, index=index)
 
-    returned_schema = module_under_test.dataframe_to_bq_schema(dataframe, [])
+    with pytest.warns(FutureWarning, match="pandas-gbq"):
+        returned_schema = module_under_test.dataframe_to_bq_schema(dataframe, [])
 
     expected_schema = (
         schema.SchemaField("str_index", "STRING", "NULLABLE"),
@@ -1334,7 +1357,9 @@ def test_dataframe_to_bq_schema_w_multiindex(module_under_test):
 
 
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
-def test_dataframe_to_bq_schema_w_bq_schema(module_under_test):
+def test_dataframe_to_bq_schema_w_bq_schema(module_under_test, monkeypatch):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     df_data = collections.OrderedDict(
         [
             ("str_column", ["hello", "world"]),
@@ -1349,7 +1374,10 @@ def test_dataframe_to_bq_schema_w_bq_schema(module_under_test):
         {"name": "bool_column", "type": "BOOL", "mode": "REQUIRED"},
     ]
 
-    returned_schema = module_under_test.dataframe_to_bq_schema(dataframe, dict_schema)
+    with pytest.warns(FutureWarning, match="pandas-gbq"):
+        returned_schema = module_under_test.dataframe_to_bq_schema(
+            dataframe, dict_schema
+        )
 
     expected_schema = (
         schema.SchemaField("str_column", "STRING", "NULLABLE"),
@@ -1360,7 +1388,11 @@ def test_dataframe_to_bq_schema_w_bq_schema(module_under_test):
 
 
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
-def test_dataframe_to_bq_schema_fallback_needed_wo_pyarrow(module_under_test):
+def test_dataframe_to_bq_schema_fallback_needed_wo_pyarrow(
+    module_under_test, monkeypatch
+):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     dataframe = pandas.DataFrame(
         data=[
             {"id": 10, "status": "FOO", "execution_date": datetime.date(2019, 5, 10)},
@@ -1388,7 +1420,11 @@ def test_dataframe_to_bq_schema_fallback_needed_wo_pyarrow(module_under_test):
 
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
 @pytest.mark.skipif(isinstance(pyarrow, mock.Mock), reason="Requires `pyarrow`")
-def test_dataframe_to_bq_schema_fallback_needed_w_pyarrow(module_under_test):
+def test_dataframe_to_bq_schema_fallback_needed_w_pyarrow(
+    module_under_test, monkeypatch
+):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     dataframe = pandas.DataFrame(
         data=[
             {"id": 10, "status": "FOO", "created_at": datetime.date(2019, 5, 10)},
@@ -1418,7 +1454,9 @@ def test_dataframe_to_bq_schema_fallback_needed_w_pyarrow(module_under_test):
 
 @pytest.mark.skipif(pandas is None, reason="Requires `pandas`")
 @pytest.mark.skipif(isinstance(pyarrow, mock.Mock), reason="Requires `pyarrow`")
-def test_dataframe_to_bq_schema_pyarrow_fallback_fails(module_under_test):
+def test_dataframe_to_bq_schema_pyarrow_fallback_fails(module_under_test, monkeypatch):
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     dataframe = pandas.DataFrame(
         data=[
             {"struct_field": {"one": 2}, "status": "FOO"},
@@ -1442,9 +1480,11 @@ def test_dataframe_to_bq_schema_pyarrow_fallback_fails(module_under_test):
 
 
 @pytest.mark.skipif(geopandas is None, reason="Requires `geopandas`")
-def test_dataframe_to_bq_schema_geography(module_under_test):
+def test_dataframe_to_bq_schema_geography(module_under_test, monkeypatch):
     from shapely import wkt
 
+    monkeypatch.setattr(module_under_test, "pandas_gbq", None)
+
     df = geopandas.GeoDataFrame(
         pandas.DataFrame(
             dict(
@@ -1455,7 +1495,10 @@ def test_dataframe_to_bq_schema_geography(module_under_test):
         ),
         geometry="geo1",
     )
-    bq_schema = module_under_test.dataframe_to_bq_schema(df, [])
+
+    with pytest.warns(FutureWarning, match="pandas-gbq"):
+        bq_schema = module_under_test.dataframe_to_bq_schema(df, [])
+
     assert bq_schema == (
         schema.SchemaField("name", "STRING"),
         schema.SchemaField("geo1", "GEOGRAPHY"),

tests/unit/test_client.py

@@ -8381,8 +8381,12 @@ def test_load_table_from_dataframe_w_automatic_schema_detection_fails(self):
             autospec=True,
             side_effect=google.api_core.exceptions.NotFound("Table not found"),
         )
+        pandas_gbq_patch = mock.patch(
+            "google.cloud.bigquery._pandas_helpers.pandas_gbq",
+            new=None,
+        )
 
-        with load_patch as load_table_from_file, get_table_patch:
+        with load_patch as load_table_from_file, get_table_patch, pandas_gbq_patch:
             with warnings.catch_warnings(record=True) as warned:
                 client.load_table_from_dataframe(
                     dataframe, self.TABLE_REF, location=self.LOCATION
@@ -8438,7 +8442,6 @@ def test_load_table_from_dataframe_w_index_and_auto_schema(self):
         load_patch = mock.patch(
             "google.cloud.bigquery.client.Client.load_table_from_file", autospec=True
         )
-
         get_table_patch = mock.patch(
             "google.cloud.bigquery.client.Client.get_table",
             autospec=True,
@@ -8450,6 +8453,7 @@ def test_load_table_from_dataframe_w_index_and_auto_schema(self):
                 ]
             ),
         )
+
         with load_patch as load_table_from_file, get_table_patch:
             client.load_table_from_dataframe(
                 dataframe, self.TABLE_REF, location=self.LOCATION
@@ -8570,10 +8574,10 @@ def test_load_table_from_dataframe_w_nullable_int64_datatype_automatic_schema(se
 
         client = self._make_client()
         dataframe = pandas.DataFrame({"x": [1, 2, None, 4]}, dtype="Int64")
+
         load_patch = mock.patch(
             "google.cloud.bigquery.client.Client.load_table_from_file", autospec=True
         )
-
         get_table_patch = mock.patch(
             "google.cloud.bigquery.client.Client.get_table",
             autospec=True,
@@ -8602,8 +8606,11 @@ def test_load_table_from_dataframe_w_nullable_int64_datatype_automatic_schema(se
 
         sent_config = load_table_from_file.mock_calls[0][2]["job_config"]
         assert sent_config.source_format == job.SourceFormat.PARQUET
-        assert tuple(sent_config.schema) == (
-            SchemaField("x", "INT64", "NULLABLE", None),
+        assert (
+            # Accept either the GoogleSQL or legacy SQL type name from pandas-gbq.
+            tuple(sent_config.schema) == (SchemaField("x", "INT64", "NULLABLE", None),)
+            or tuple(sent_config.schema)
+            == (SchemaField("x", "INTEGER", "NULLABLE", None),)
         )
 
     def test_load_table_from_dataframe_struct_fields(self):
@@ -8749,14 +8756,22 @@ def test_load_table_from_dataframe_array_fields_w_auto_schema(self):
             data=records, columns=["float_column", "array_column"]
         )
 
-        expected_schema = [
+        expected_schema_googlesql = [
             SchemaField("float_column", "FLOAT"),
             SchemaField(
                 "array_column",
                 "INT64",
                 mode="REPEATED",
             ),
         ]
+        expected_schema_legacy_sql = [
+            SchemaField("float_column", "FLOAT"),
+            SchemaField(
+                "array_column",
+                "INTEGER",
+                mode="REPEATED",
+            ),
+        ]
 
         load_patch = mock.patch(
             "google.cloud.bigquery.client.Client.load_table_from_file", autospec=True
@@ -8792,7 +8807,10 @@ def test_load_table_from_dataframe_array_fields_w_auto_schema(self):
 
         sent_config = load_table_from_file.mock_calls[0][2]["job_config"]
         assert sent_config.source_format == job.SourceFormat.PARQUET
-        assert sent_config.schema == expected_schema
+        assert (
+            sent_config.schema == expected_schema_googlesql
+            or sent_config.schema == expected_schema_legacy_sql
+        )
 
     def test_load_table_from_dataframe_w_partial_schema(self):
         pandas = pytest.importorskip("pandas")