Backport PR pandas-dev#45257 CI: Simplify GHA dependency files (pandas-dev#45257)

Author: mroeschke

.github/workflows/posix.yml

@@ -28,13 +28,13 @@ jobs:
           [actions-38-downstream_compat.yaml, "not slow and not network and not clipboard", "", "", "", "", ""],
           [actions-38-minimum_versions.yaml, "slow", "", "", "", "", ""],
           [actions-38-minimum_versions.yaml, "not slow and not network and not clipboard", "", "", "", "", ""],
-          [actions-38-locale_slow.yaml, "slow", "language-pack-it xsel", "it_IT.utf8", "it_IT.utf8", "", ""],
+          [actions-38.yaml, "not slow and not network", "language-pack-it xsel", "it_IT.utf8", "it_IT.utf8", "", ""],
+          [actions-38.yaml, "not slow and not network", "language-pack-zh-hans xsel", "zh_CN.utf8", "zh_CN.utf8", "", ""],
           [actions-38.yaml, "not slow and not clipboard", "", "", "", "", ""],
-          [actions-38-slow.yaml, "slow", "", "", "", "", ""],
-          [actions-38-locale.yaml, "not slow and not network", "language-pack-zh-hans xsel", "zh_CN.utf8", "zh_CN.utf8", "", ""],
-          [actions-39-slow.yaml, "slow", "", "", "", "", ""],
+          [actions-38.yaml, "slow", "", "", "", "", ""],
           [actions-pypy-38.yaml, "not slow and not clipboard", "", "", "", "", "--max-worker-restart 0"],
           [actions-39-numpydev.yaml, "not slow and not network", "xsel", "", "", "deprecate", "-W error"],
+          [actions-39.yaml, "slow", "", "", "", "", ""],
           [actions-39.yaml, "not slow and not clipboard", "", "", "", "", ""]
         ]
       fail-fast: false

ci/deps/actions-38-locale.yaml

@@ -1,20 +1,50 @@
 name: pandas-dev
 channels:
-  - defaults
   - conda-forge
 dependencies:
   - python=3.8
 
-  # tools
-  - cython>=0.29.24
+  # test dependencies
+  - cython=0.29.24
   - pytest>=6.0
   - pytest-cov
   - pytest-xdist>=1.31
   - hypothesis>=5.5.3
+  - psutil
 
-  # pandas dependencies
-  - numpy
+  # required dependencies
   - python-dateutil
-  - nomkl
+  - numpy
   - pytz
+
+  # optional dependencies
+  - beautifulsoup4
+  - blosc
+  - bottleneck
+  - fastparquet
+  - fsspec
+  - html5lib
+  - gcsfs
+  - jinja2
+  - lxml
+  - matplotlib
+  - numba
+  - numexpr
+  - openpyxl
+  - odfpy
+  - pandas-gbq
+  - psycopg2
+  - pymysql
+  - pytables
+  - pyarrow
+  - pyreadstat
+  - pyxlsb
+  - s3fs
+  - scipy
+  - sqlalchemy
   - tabulate
+  - xarray
+  - xlrd
+  - xlsxwriter
+  - xlwt
+  - zstandard

ci/deps/actions-38-locale_slow.yaml

@@ -1,12 +1,7 @@
-import locale
-
 import numpy as np
 import pytest
 
-from pandas.compat import (
-    is_platform_windows,
-    np_version_under1p20,
-)
+from pandas.compat import np_version_under1p20
 
 import pandas as pd
 import pandas._testing as tm
@@ -47,7 +42,7 @@ def test_floating_array_constructor():
         FloatingArray(values)
 
 
-def test_floating_array_disallows_float16(request):
+def test_floating_array_disallows_float16():
     # GH#44715
     arr = np.array([1, 2], dtype=np.float16)
     mask = np.array([False, False])
@@ -56,11 +51,10 @@ def test_floating_array_disallows_float16(request):
     with pytest.raises(TypeError, match=msg):
         FloatingArray(arr, mask)
 
-    if np_version_under1p20 or (
-        locale.getlocale()[0] != "en_US" and not is_platform_windows()
-    ):
-        # the locale condition may need to be refined; this fails on
-        #  the CI in the ZH_CN build
+
+def test_floating_array_disallows_Float16_dtype(request):
+    # GH#44715
+    if np_version_under1p20:
         # https://github.com/numpy/numpy/issues/20512
         mark = pytest.mark.xfail(reason="numpy does not raise on np.dtype('Float16')")
         request.node.add_marker(mark)

ci/deps/actions-38-slow.yaml

@@ -1,6 +1,6 @@
 import pytest
 
-from pandas.compat import pa_version_under4p0
+from pandas.compat import pa_version_under3p0
 
 from pandas.core.dtypes.dtypes import PeriodDtype
 
@@ -72,7 +72,7 @@ def test_arrow_array_missing():
 
 
 @pytest.mark.xfail(
-    pa_version_under4p0, reason="pyarrow incorrectly uses pandas internals API"
+    pa_version_under3p0, reason="pyarrow incorrectly uses pandas internals API"
 )
 def test_arrow_table_roundtrip():
     from pandas.core.arrays._arrow_utils import ArrowPeriodType
@@ -94,7 +94,7 @@ def test_arrow_table_roundtrip():
 
 
 @pytest.mark.xfail(
-    pa_version_under4p0, reason="pyarrow incorrectly uses pandas internals API"
+    pa_version_under3p0, reason="pyarrow incorrectly uses pandas internals API"
 )
 def test_arrow_load_from_zero_chunks():
     # GH-41040
@@ -115,7 +115,7 @@ def test_arrow_load_from_zero_chunks():
 
 
 @pytest.mark.xfail(
-    pa_version_under4p0, reason="pyarrow incorrectly uses pandas internals API"
+    pa_version_under3p0, reason="pyarrow incorrectly uses pandas internals API"
 )
 def test_arrow_table_roundtrip_without_metadata():
     arr = PeriodArray([1, 2, 3], freq="H")

ci/deps/actions-38.yaml

@@ -738,10 +738,9 @@ def test_bad_sheetname_raises(self, read_ext, sheet_name):
 
     def test_missing_file_raises(self, read_ext):
         bad_file = f"foo{read_ext}"
-        # CI tests with zh_CN.utf8, translates to "No such file or directory"
-        with pytest.raises(
-            FileNotFoundError, match=r"(No such file or directory|没有那个文件或目录)"
-        ):
+        # CI tests with other languages, translates to "No such file or directory"
+        match = r"(No such file or directory|没有那个文件或目录|File o directory non esistente)"
+        with pytest.raises(FileNotFoundError, match=match):
             pd.read_excel(bad_file)
 
     def test_corrupt_bytes_raises(self, read_ext, engine):

ci/deps/actions-39-slow.yaml

@@ -2,7 +2,7 @@
 import numpy as np
 import pytest
 
-from pandas.compat.pyarrow import pa_version_under4p0
+from pandas.compat.pyarrow import pa_version_under3p0
 
 import pandas as pd
 import pandas._testing as tm
@@ -87,7 +87,7 @@ def test_basic(self):
                 ),
             }
         )
-        if not pa_version_under4p0:
+        if not pa_version_under3p0:
             # older pyarrow incorrectly uses pandas internal API, so
             #  constructs invalid Block
             df["periods"] = pd.period_range("2013", freq="M", periods=3)

pandas/tests/arrays/floating/test_construction.py

@@ -15,7 +15,7 @@
 
 from pandas.compat.pyarrow import (
     pa_version_under2p0,
-    pa_version_under4p0,
+    pa_version_under3p0,
     pa_version_under5p0,
     pa_version_under6p0,
 )
@@ -654,8 +654,10 @@ def test_use_nullable_dtypes(self, engine, request):
             "float",
             pytest.param(
                 "period[D]",
+                # Note: I don't know exactly what version the cutoff is;
+                #  On the CI it fails with 1.0.1 & 2.0.0
                 marks=pytest.mark.xfail(
-                    pa_version_under4p0,
+                    pa_version_under3p0,
                     reason="pyarrow uses pandas internal API incorrectly",
                 ),
             ),
@@ -898,7 +900,7 @@ def test_pyarrow_backed_string_array(self, pa, string_storage):
 
     @td.skip_if_no("pyarrow")
     @pytest.mark.xfail(
-        pa_version_under4p0, reason="pyarrow uses pandas internal API incorrectly"
+        pa_version_under3p0, reason="pyarrow uses pandas internal API incorrectly"
     )
     def test_additional_extension_types(self, pa):
         # test additional ExtensionArrays that are supported through the

pandas/tests/arrays/period/test_arrow_compat.py

@@ -288,26 +288,26 @@ def test_to_datetime_format_microsecond(self, cache):
                 "%m/%d/%Y %I:%M %p",
                 Timestamp("2010-01-10 20:14"),
                 marks=pytest.mark.xfail(
-                    locale.getlocale()[0] == "zh_CN",
-                    reason="fail on a CI build with LC_ALL=zh_CN.utf8",
+                    locale.getlocale()[0] in ("zh_CN", "it_IT"),
+                    reason="fail on a CI build with LC_ALL=zh_CN.utf8/it_IT.utf8",
                 ),
             ),
             pytest.param(
                 "01/10/2010 07:40 AM",
                 "%m/%d/%Y %I:%M %p",
                 Timestamp("2010-01-10 07:40"),
                 marks=pytest.mark.xfail(
-                    locale.getlocale()[0] == "zh_CN",
-                    reason="fail on a CI build with LC_ALL=zh_CN.utf8",
+                    locale.getlocale()[0] in ("zh_CN", "it_IT"),
+                    reason="fail on a CI build with LC_ALL=zh_CN.utf8/it_IT.utf8",
                 ),
             ),
             pytest.param(
                 "01/10/2010 09:12:56 AM",
                 "%m/%d/%Y %I:%M:%S %p",
                 Timestamp("2010-01-10 09:12:56"),
                 marks=pytest.mark.xfail(
-                    locale.getlocale()[0] == "zh_CN",
-                    reason="fail on a CI build with LC_ALL=zh_CN.utf8",
+                    locale.getlocale()[0] in ("zh_CN", "it_IT"),
+                    reason="fail on a CI build with LC_ALL=zh_CN.utf8/it_IT.utf8",
                 ),
             ),
         ],