Testing: Pull makeTimeDataFrame and makeMixedDataFrame from pueblo

amotl · amotl · commit db7ba4d0e1f4 · 2024-01-30T21:06:38.000+01:00
`pandas._testing.{makeTimeDataFrame,makeMixedDataFrame}` were removed on
behalf of pandas 2.2.0. This patch pulls corresponding polyfills from
a hostel package.
diff --git a/docs/by-example/sqlalchemy/dataframe.rst b/docs/by-example/sqlalchemy/dataframe.rst
@@ -76,8 +76,8 @@ The package provides a ``bulk_insert`` function to use the
 workload across multiple batches, using a defined chunk size.
 
     >>> import sqlalchemy as sa
-    >>> from pandas._testing import makeTimeDataFrame
     >>> from crate.client.sqlalchemy.support import insert_bulk
+    >>> from pueblo.testing.pandas import makeTimeDataFrame
     ...
     >>> # Define number of records, and chunk size.
     >>> INSERT_RECORDS = 42
@@ -159,8 +159,8 @@ in a batched/chunked manner, using a defined chunk size, effectively using the
 pandas implementation introduced in the previous section.
 
     >>> import dask.dataframe as dd
-    >>> from pandas._testing import makeTimeDataFrame
     >>> from crate.client.sqlalchemy.support import insert_bulk
+    >>> from pueblo.testing.pandas import makeTimeDataFrame
     ...
     >>> # Define the number of records, the number of computing partitions,
     >>> # and the chunk size of each database insert operation.
diff --git a/setup.py b/setup.py
@@ -76,6 +76,7 @@ def read(path):
               'stopit>=1.1.2,<2',
               'flake8>=4,<8',
               'pandas<2.3',
+              'pueblo>=0.0.7',
               'pytz',
               ],
         doc=['sphinx>=3.5,<8',
diff --git a/src/crate/client/sqlalchemy/tests/bulk_test.py b/src/crate/client/sqlalchemy/tests/bulk_test.py
@@ -176,8 +176,8 @@ def test_bulk_save_pandas(self, mock_cursor):
         """
         Verify bulk INSERT with pandas.
         """
-        from pandas._testing import makeTimeDataFrame
         from crate.client.sqlalchemy.support import insert_bulk
+        from pueblo.testing.pandas import makeTimeDataFrame
 
         # 42 records / 8 chunksize = 5.25, which means 6 batches will be emitted.
         INSERT_RECORDS = 42
@@ -216,8 +216,8 @@ def test_bulk_save_dask(self, mock_cursor):
         Verify bulk INSERT with Dask.
         """
         import dask.dataframe as dd
-        from pandas._testing import makeTimeDataFrame
         from crate.client.sqlalchemy.support import insert_bulk
+        from pueblo.testing.pandas import makeTimeDataFrame
 
         # 42 records / 4 partitions means each partition has a size of 10.5 elements.
         # Because the chunk size 8 is slightly smaller than 10, the partition will not
