Merge branch '1.4.x' of https://github.com/pandas-dev/pandas into 1.4.x

Author: jbrockmendel

.github/workflows/posix.yml

@@ -50,7 +50,7 @@ jobs:
       COVERAGE: ${{ !contains(matrix.settings[0], 'pypy') }}
     concurrency:
       # https://github.202132.xyzmunity/t/concurrecy-not-work-for-push/183068/7
-      group: ${{ github.event_name == 'push' && github.run_number || github.ref }}-${{ matrix.settings[0] }}-${{ matrix.settings[1] }}
+      group: ${{ github.event_name == 'push' && github.run_number || github.ref }}-${{ matrix.settings[0] }}-${{ matrix.settings[1] }}-${{ matrix.settings[2] }}
       cancel-in-progress: true
 
     services:

ci/deps/actions-310.yaml

@@ -2,7 +2,7 @@ name: pandas-dev
 channels:
   - conda-forge
 dependencies:
-  - python=3.9
+  - python=3.10
 
   # test dependencies
   - cython=0.29.24

doc/source/whatsnew/v1.4.1.rst

@@ -18,17 +18,18 @@ Fixed regressions
 - Regression in :func:`.assert_frame_equal` not respecting ``check_flags=False`` (:issue:`45554`)
 - Regression in :meth:`Series.fillna` with ``downcast=False`` incorrectly downcasting ``object`` dtype (:issue:`45603`)
 - Regression in :meth:`DataFrame.iat` setting values leading to not propagating correctly in subsequent lookups (:issue:`45684`)
-- Regression in :meth:`DataFrame.loc.__setitem__` losing :class:`Index` name if :class:`DataFrame` was empty before (:issue:`45621`)
-- Regression in :func:`join` with overlapping :class:`IntervalIndex` raising an ``InvalidIndexError`` (:issue:`45661`)
-
+- Regression when setting values with :meth:`DataFrame.loc` losing :class:`Index` name if :class:`DataFrame` was empty before (:issue:`45621`)
+- Regression in :meth:`~Index.join` with overlapping :class:`IntervalIndex` raising an ``InvalidIndexError`` (:issue:`45661`)
+- Regression in :func:`read_sql` with a DBAPI2 connection that is not an instance of ``sqlite3.Connection`` incorrectly requiring SQLAlchemy be installed (:issue:`45660`)
+-
 
 .. ---------------------------------------------------------------------------
 
 .. _whatsnew_141.bug_fixes:
 
 Bug fixes
 ~~~~~~~~~
-- Fixed segfault in :meth:``DataFrame.to_json`` when dumping tz-aware datetimes in Python 3.10 (:issue:`42130`)
+- Fixed segfault in :meth:`DataFrame.to_json` when dumping tz-aware datetimes in Python 3.10 (:issue:`42130`)
 - Stopped emitting unnecessary ``FutureWarning`` in :meth:`DataFrame.sort_values` with sparse columns (:issue:`45618`)
 - Fixed window aggregations in :meth:`DataFrame.rolling` and :meth:`Series.rolling` to skip over unused elements (:issue:`45647`)
 - Bug in :func:`api.types.is_bool_dtype` was raising an ``AttributeError`` when evaluating a categorical :class:`Series` (:issue:`45615`)

environment.yml

@@ -35,6 +35,7 @@ dependencies:
   - sphinx
   - sphinx-panels
   - numpydoc < 1.2  # 2021-02-09 1.2dev breaking CI
+  - pydata-sphinx-theme
   - types-python-dateutil
   - types-PyMySQL
   - types-pytz
@@ -120,6 +121,5 @@ dependencies:
   - tabulate>=0.8.3  # DataFrame.to_markdown
   - natsort  # DataFrame.sort_values
   - pip:
-    - pydata-sphinx-theme
     - pandas-dev-flaker==0.2.0
     - pytest-cython

pandas/io/sql.py

@@ -747,12 +747,15 @@ def pandasSQL_builder(con, schema: str | None = None):
     if isinstance(con, sqlite3.Connection) or con is None:
         return SQLiteDatabase(con)
 
-    sqlalchemy = import_optional_dependency("sqlalchemy")
+    sqlalchemy = import_optional_dependency("sqlalchemy", errors="ignore")
 
     if isinstance(con, str):
-        con = sqlalchemy.create_engine(con)
+        if sqlalchemy is None:
+            raise ImportError("Using URI string without sqlalchemy installed.")
+        else:
+            con = sqlalchemy.create_engine(con)
 
-    if isinstance(con, sqlalchemy.engine.Connectable):
+    if sqlalchemy is not None and isinstance(con, sqlalchemy.engine.Connectable):
         return SQLDatabase(con, schema=schema)
 
     warnings.warn(

pandas/tests/base/test_unique.py

@@ -105,6 +105,9 @@ def test_nunique_null(null_obj, index_or_series_obj):
 
 
 @pytest.mark.single
+@pytest.mark.xfail(
+    reason="Flaky in the CI. Remove once CI has a single build: GH 44584", strict=False
+)
 def test_unique_bad_unicode(index_or_series):
     # regression test for #34550
     uval = "\ud83d"  # smiley emoji

pandas/tests/io/parser/test_network.py

@@ -7,6 +7,7 @@
     StringIO,
 )
 import logging
+import os
 
 import numpy as np
 import pytest
@@ -256,6 +257,12 @@ def test_read_csv_handles_boto_s3_object(self, s3_resource, tips_file):
         expected = read_csv(tips_file)
         tm.assert_frame_equal(result, expected)
 
+    @pytest.mark.skipif(
+        os.environ.get("PANDAS_CI", "0") == "1",
+        reason="This test can hang in our CI min_versions build "
+        "and leads to '##[error]The runner has "
+        "received a shutdown signal...' in GHA. GH: 45651",
+    )
     def test_read_csv_chunked_download(self, s3_resource, caplog, s3so):
         # 8 MB, S3FS uses 5MB chunks
         import s3fs

pandas/tests/io/test_sql.py

@@ -1534,9 +1534,25 @@ def test_sql_open_close(self, test_frame3):
     @pytest.mark.skipif(SQLALCHEMY_INSTALLED, reason="SQLAlchemy is installed")
     def test_con_string_import_error(self):
         conn = "mysql://root@localhost/pandas"
-        with pytest.raises(ImportError, match="SQLAlchemy"):
+        msg = "Using URI string without sqlalchemy installed"
+        with pytest.raises(ImportError, match=msg):
             sql.read_sql("SELECT * FROM iris", conn)
 
+    @pytest.mark.skipif(SQLALCHEMY_INSTALLED, reason="SQLAlchemy is installed")
+    def test_con_unknown_dbapi2_class_does_not_error_without_sql_alchemy_installed(
+        self,
+    ):
+        class MockSqliteConnection:
+            def __init__(self, *args, **kwargs):
+                self.conn = sqlite3.Connection(*args, **kwargs)
+
+            def __getattr__(self, name):
+                return getattr(self.conn, name)
+
+        conn = MockSqliteConnection(":memory:")
+        with tm.assert_produces_warning(UserWarning):
+            sql.read_sql("SELECT 1", conn)
+
     def test_read_sql_delegate(self):
         iris_frame1 = sql.read_sql_query("SELECT * FROM iris", self.conn)
         iris_frame2 = sql.read_sql("SELECT * FROM iris", self.conn)

pandas/tests/io/test_user_agent.py

@@ -5,6 +5,7 @@
 import http.server
 from io import BytesIO
 import multiprocessing
+import os
 import socket
 import time
 import urllib.error
@@ -16,6 +17,13 @@
 import pandas as pd
 import pandas._testing as tm
 
+pytestmark = pytest.mark.skipif(
+    os.environ.get("PANDAS_CI", "0") == "1",
+    reason="This test can hang in our CI min_versions build "
+    "and leads to '##[error]The runner has "
+    "received a shutdown signal...' in GHA. GH 45651",
+)
+
 
 class BaseUserAgentResponder(http.server.BaseHTTPRequestHandler):
     """
@@ -242,7 +250,9 @@ def responder(request):
             pd.read_parquet,
             "fastparquet",
             # TODO(ArrayManager) fastparquet
-            marks=td.skip_array_manager_not_yet_implemented,
+            marks=[
+                td.skip_array_manager_not_yet_implemented,
+            ],
         ),
         (PickleUserAgentResponder, pd.read_pickle, None),
         (StataUserAgentResponder, pd.read_stata, None),
@@ -277,7 +287,9 @@ def test_server_and_default_headers(responder, read_method, parquet_engine):
             pd.read_parquet,
             "fastparquet",
             # TODO(ArrayManager) fastparquet
-            marks=td.skip_array_manager_not_yet_implemented,
+            marks=[
+                td.skip_array_manager_not_yet_implemented,
+            ],
         ),
         (PickleUserAgentResponder, pd.read_pickle, None),
         (StataUserAgentResponder, pd.read_stata, None),

requirements-dev.txt

@@ -21,6 +21,7 @@ gitdb
 sphinx
 sphinx-panels
 numpydoc < 1.2
+pydata-sphinx-theme
 types-python-dateutil
 types-PyMySQL
 types-pytz
@@ -83,7 +84,6 @@ cftime
 pyreadstat
 tabulate>=0.8.3
 natsort
-pydata-sphinx-theme
 pandas-dev-flaker==0.2.0
 pytest-cython
 setuptools>=51.0.0