CI: debug azure timeouts

.github/workflows/database.yml

@@ -87,7 +87,7 @@ jobs:
       uses: ./.github/actions/build_pandas
 
     - name: Test
-      run: pytest -m "${{ env.PATTERN }}" -n 2 --dist=loadfile --cov=pandas --cov-report=xml pandas/tests/io
+      run: pytest -m "${{ env.PATTERN }}" -n 2 --dist=loadfile --cov=pandas --cov-report=xml pandas/tests/io --timeout=300
       if: always()
 
     - name: Build Version

.github/workflows/python-dev.yml

@@ -45,7 +45,7 @@ jobs:
         pip install -i https://pypi.anaconda.org/scipy-wheels-nightly/simple numpy
         pip install git+https://github.com/pytest-dev/pytest.git
         pip install git+https://github.com/nedbat/coveragepy.git
-        pip install cython python-dateutil pytz hypothesis pytest-xdist pytest-cov
+        pip install cython python-dateutil pytz hypothesis pytest-xdist pytest-cov pytest-timeout
         pip list
 
     - name: Build Pandas

ci/azure/posix.yml

@@ -12,12 +12,12 @@ jobs:
         py38_macos_1:
           ENV_FILE: ci/deps/azure-macos-38.yaml
           CONDA_PY: "38"
-          PATTERN: "not slow and not network"
+          PATTERN: "not slow and not network and not single"
           PYTEST_TARGET: "pandas/tests/[a-h]*"
         py38_macos_2:
           ENV_FILE: ci/deps/azure-macos-38.yaml
           CONDA_PY: "38"
-          PATTERN: "not slow and not network"
+          PATTERN: "not slow and not network and not single"
           PYTEST_TARGET: "pandas/tests/[i-z]*"
 
   steps:

ci/deps/actions-38-db-min.yaml

@@ -45,3 +45,5 @@ dependencies:
   - psycopg2=2.8.4
   - pymysql=0.10.1
   - sqlalchemy=1.3.11
+  - pip:
+      - pytest-timeout

ci/deps/actions-38-db.yaml

@@ -51,3 +51,4 @@ dependencies:
     - coverage
     - pandas-datareader
     - pyxlsb
+    - pytest-timeout

ci/deps/actions-38-locale.yaml

@@ -39,3 +39,4 @@ dependencies:
   - pip
   - pip:
     - pyxlsb
+    - pytest-timeout

ci/deps/actions-38-locale_slow.yaml

@@ -28,3 +28,5 @@ dependencies:
   - xlsxwriter=1.2.2
   - xlwt=1.3.0
   - html5lib=1.1
+  - pip:
+      - pytest-timeout

ci/deps/actions-38-minimum_versions.yaml

@@ -29,3 +29,5 @@ dependencies:
   - xlsxwriter=1.2.2
   - xlwt=1.3.0
   - html5lib=1.1
+  - pip:
+      - pytest-timeout

ci/deps/actions-38-slow.yaml

@@ -36,3 +36,5 @@ dependencies:
   - moto
   - flask
   - numba
+  - pip:
+      - pytest-timeout

ci/deps/actions-38.yaml

@@ -18,3 +18,5 @@ dependencies:
   - nomkl
   - pytz
   - tabulate==0.8.7
+  - pip:
+      - pytest-timeout

ci/deps/actions-39-numpydev.yaml

@@ -20,3 +20,4 @@ dependencies:
     - "--pre"
     - "numpy"
     - "scipy"
+    - pytest-timeout

ci/deps/actions-39-slow.yaml

@@ -41,3 +41,4 @@ dependencies:
   - pip
   - pip:
     - pyxlsb
+    - pytest-timeout

ci/deps/actions-39.yaml

@@ -40,3 +40,4 @@ dependencies:
   - pip
   - pip:
     - pyxlsb
+    - pytest-timeout

ci/deps/azure-macos-38.yaml

@@ -35,3 +35,4 @@ dependencies:
     - cython>=0.29.21
     - pyreadstat
     - pyxlsb
+    - pytest-timeout

ci/deps/azure-windows-38.yaml

@@ -34,3 +34,5 @@ dependencies:
   - xlrd
   - xlsxwriter
   - xlwt
+  - pip:
+      - pytest-timeout

ci/deps/azure-windows-39.yaml

@@ -41,3 +41,4 @@ dependencies:
   - pip
   - pip:
     - pyxlsb
+    - pytest-timeout

ci/run_tests.sh

@@ -19,7 +19,7 @@ if [[ $(uname) == "Linux" && -z $DISPLAY ]]; then
     XVFB="xvfb-run "
 fi
 
-PYTEST_CMD="${XVFB}pytest -m \"$PATTERN\" -n $PYTEST_WORKERS --dist=loadfile $TEST_ARGS $COVERAGE $PYTEST_TARGET"
+PYTEST_CMD="${XVFB}pytest -m \"$PATTERN\" -n $PYTEST_WORKERS --dist=loadfile $TEST_ARGS $COVERAGE $PYTEST_TARGET --timeout=60"
 
 if [[ $(uname) != "Linux"  && $(uname) != "Darwin" ]]; then
     # GH#37455 windows py38 build appears to be running out of memory

environment.yml

@@ -63,6 +63,7 @@ dependencies:
   - pytest-xdist>=1.31
   - pytest-asyncio
   - pytest-instafail
+  - pytest-timeout
 
   # downstream tests
   - seaborn

pandas/tests/io/parser/common/test_index.py

@@ -16,6 +16,7 @@
 import pandas._testing as tm
 
 xfail_pyarrow = pytest.mark.usefixtures("pyarrow_xfail")
+skip_pyarrow = pytest.mark.usefixtures("pyarrow_skip")
 
 
 @pytest.mark.parametrize(
@@ -271,7 +272,7 @@ def test_empty_with_index(all_parsers):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 def test_empty_with_multi_index(all_parsers):
     # see gh-10467
     data = "x,y,z"
@@ -284,7 +285,7 @@ def test_empty_with_multi_index(all_parsers):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 def test_empty_with_reversed_multi_index(all_parsers):
     data = "x,y,z"
     parser = all_parsers

pandas/tests/io/parser/common/test_ints.py

@@ -13,7 +13,6 @@
 )
 import pandas._testing as tm
 
-xfail_pyarrow = pytest.mark.usefixtures("pyarrow_xfail")
 skip_pyarrow = pytest.mark.usefixtures("pyarrow_skip")
 
 
@@ -119,7 +118,7 @@ def test_int64_min_issues(all_parsers):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize("conv", [None, np.int64, np.uint64])
 def test_int64_overflow(all_parsers, conv):
     data = """ID
@@ -163,7 +162,7 @@ def test_int64_overflow(all_parsers, conv):
             parser.read_csv(StringIO(data), converters={"ID": conv})
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize(
     "val", [np.iinfo(np.uint64).max, np.iinfo(np.int64).max, np.iinfo(np.int64).min]
 )
@@ -177,7 +176,7 @@ def test_int64_uint64_range(all_parsers, val):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize(
     "val", [np.iinfo(np.uint64).max + 1, np.iinfo(np.int64).min - 1]
 )
@@ -191,7 +190,7 @@ def test_outside_int64_uint64_range(all_parsers, val):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize("exp_data", [[str(-1), str(2 ** 63)], [str(2 ** 63), str(-1)]])
 def test_numeric_range_too_wide(all_parsers, exp_data):
     # No numerical dtype can hold both negative and uint64

pandas/tests/io/parser/conftest.py

@@ -82,10 +82,10 @@ def csv1(datapath):
     return os.path.join(datapath("io", "data", "csv"), "test1.csv")
 
 
-_cParserHighMemory = CParserHighMemory()
-_cParserLowMemory = CParserLowMemory()
-_pythonParser = PythonParser()
-_pyarrowParser = PyArrowParser()
+_cParserHighMemory = CParserHighMemory
+_cParserLowMemory = CParserLowMemory
+_pythonParser = PythonParser
+_pyarrowParser = PyArrowParser
 
 _py_parsers_only = [_pythonParser]
 _c_parsers_only = [_cParserHighMemory, _cParserLowMemory]
@@ -105,37 +105,38 @@ def all_parsers(request):
     """
     Fixture all of the CSV parsers.
     """
-    if request.param.engine == "pyarrow":
+    parser = request.param()
+    if parser.engine == "pyarrow":
         pytest.importorskip("pyarrow", VERSIONS["pyarrow"])
         # Try setting num cpus to 1 to avoid hangs?
         import pyarrow
 
         pyarrow.set_cpu_count(1)
-    return request.param
+    return parser
 
 
 @pytest.fixture(params=_c_parsers_only, ids=_c_parser_ids)
 def c_parser_only(request):
     """
     Fixture all of the CSV parsers using the C engine.
     """
-    return request.param
+    return request.param()
 
 
 @pytest.fixture(params=_py_parsers_only, ids=_py_parser_ids)
 def python_parser_only(request):
     """
     Fixture all of the CSV parsers using the Python engine.
     """
-    return request.param
+    return request.param()
 
 
 @pytest.fixture(params=_pyarrow_parsers_only, ids=_pyarrow_parsers_ids)
 def pyarrow_parser_only(request):
     """
     Fixture all of the CSV parsers using the Pyarrow engine.
     """
-    return request.param
+    return request.param()
 
 
 def _get_all_parser_float_precision_combinations():
@@ -147,7 +148,7 @@ def _get_all_parser_float_precision_combinations():
     ids = []
     for parser, parser_id in zip(_all_parsers, _all_parser_ids):
         for precision in parser.float_precision_choices:
-            params.append((parser, precision))
+            params.append((parser(), precision))
             ids.append(f"{parser_id}-{precision}")
 
     return {"params": params, "ids": ids}

pandas/tests/io/parser/test_parse_dates.py

@@ -43,6 +43,7 @@
 from pandas.io.parsers import read_csv
 
 xfail_pyarrow = pytest.mark.usefixtures("pyarrow_xfail")
+skip_pyarrow = pytest.mark.usefixtures("pyarrow_skip")
 
 # constant
 _DEFAULT_DATETIME = datetime(1, 1, 1)
@@ -1573,7 +1574,7 @@ def test_parse_timezone(all_parsers):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize(
     "date_string",
     ["32/32/2019", "02/30/2019", "13/13/2019", "13/2019", "a3/11/2018", "10/11/2o17"],
@@ -1585,7 +1586,7 @@ def test_invalid_parse_delimited_date(all_parsers, date_string):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize(
     "date_string,dayfirst,expected",
     [
@@ -1608,7 +1609,7 @@ def test_parse_delimited_date_swap_no_warning(
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize(
     "date_string,dayfirst,expected",
     [
@@ -1643,6 +1644,8 @@ def _helper_hypothesis_delimited_date(call, date_string, **kwargs):
     return msg, result
 
 
+@pytest.mark.slow
+@skip_pyarrow
 @given(date_strategy)
 @settings(deadline=None)
 @pytest.mark.parametrize("delimiter", list(" -./"))
@@ -1678,7 +1681,7 @@ def test_hypothesis_delimited_date(date_format, dayfirst, delimiter, test_dateti
     assert result == expected
 
 
-@xfail_pyarrow
+@skip_pyarrow
 @pytest.mark.parametrize(
     "names, usecols, parse_dates, missing_cols",
     [
@@ -1711,7 +1714,7 @@ def test_missing_parse_dates_column_raises(
         )
 
 
-@xfail_pyarrow
+@skip_pyarrow
 def test_date_parser_and_names(all_parsers):
     # GH#33699
     parser = all_parsers
@@ -1721,7 +1724,7 @@ def test_date_parser_and_names(all_parsers):
     tm.assert_frame_equal(result, expected)
 
 
-@xfail_pyarrow
+@skip_pyarrow
 def test_date_parser_usecols_thousands(all_parsers):
     # GH#39365
     data = """A,B,C

pyproject.toml

@@ -67,6 +67,7 @@ markers = [
   "arm_slow: mark a test as slow for arm64 architecture",
   "arraymanager: mark a test to run with ArrayManager enabled",
 ]
+log_cli = 1
 
 [tool.mypy]
 platform = "linux-64"

requirements-dev.txt

@@ -41,6 +41,7 @@ pytest-cov
 pytest-xdist>=1.31
 pytest-asyncio
 pytest-instafail
+pytest-timeout
 seaborn
 statsmodels
 ipywidgets