TST: conform tests for datetutil > 2.6.1 (#18253)

Author: jreback

pandas/conftest.py

@@ -1,8 +1,10 @@
 import pytest
 
+from distutils.version import LooseVersion
 import numpy
 import pandas
 import pandas.util.testing as tm
+import dateutil
 
 
 def pytest_addoption(parser):
@@ -65,3 +67,11 @@ def ip():
     pytest.importorskip('IPython', minversion="6.0.0")
     from IPython.core.interactiveshell import InteractiveShell
     return InteractiveShell()
+
+
+is_dateutil_le_261 = pytest.mark.skipif(
+    LooseVersion(dateutil.__version__) > '2.6.1',
+    reason="dateutil api change version")
+is_dateutil_gt_261 = pytest.mark.skipif(
+    LooseVersion(dateutil.__version__) <= '2.6.1',
+    reason="dateutil stable version")

pandas/tests/indexes/datetimes/test_tools.py

@@ -12,6 +12,7 @@
 from distutils.version import LooseVersion
 
 import pandas as pd
+from pandas.conftest import is_dateutil_le_261, is_dateutil_gt_261
 from pandas._libs import tslib
 from pandas._libs.tslibs import parsing
 from pandas.core.tools import datetimes as tools
@@ -996,7 +997,7 @@ def test_dayfirst(self, cache):
 
 class TestGuessDatetimeFormat(object):
 
-    @pytest.mark.xfail(reason="GH18141 - dateutil > 2.6.1 broken")
+    @is_dateutil_le_261
     def test_guess_datetime_format_for_array(self):
         tm._skip_if_not_us_locale()
         expected_format = '%Y-%m-%d %H:%M:%S.%f'
@@ -1017,6 +1018,27 @@ def test_guess_datetime_format_for_array(self):
                 [np.nan, np.nan, np.nan], dtype='O'))
         assert format_for_string_of_nans is None
 
+    @is_dateutil_gt_261
+    def test_guess_datetime_format_for_array_gt_261(self):
+        tm._skip_if_not_us_locale()
+        expected_format = '%Y-%m-%d %H:%M:%S.%f'
+        dt_string = datetime(2011, 12, 30, 0, 0, 0).strftime(expected_format)
+
+        test_arrays = [
+            np.array([dt_string, dt_string, dt_string], dtype='O'),
+            np.array([np.nan, np.nan, dt_string], dtype='O'),
+            np.array([dt_string, 'random_string'], dtype='O'),
+        ]
+
+        for test_array in test_arrays:
+            assert tools._guess_datetime_format_for_array(
+                test_array) is None
+
+        format_for_string_of_nans = tools._guess_datetime_format_for_array(
+            np.array(
+                [np.nan, np.nan, np.nan], dtype='O'))
+        assert format_for_string_of_nans is None
+
 
 class TestToDatetimeInferFormat(object):
 

pandas/tests/scalar/test_parsing.py

@@ -3,14 +3,12 @@
 Tests for Timestamp parsing, aimed at pandas/_libs/tslibs/parsing.pyx
 """
 from datetime import datetime
-
 import numpy as np
 import pytest
 from dateutil.parser import parse
-
+from pandas.conftest import is_dateutil_le_261, is_dateutil_gt_261
 from pandas import compat
 from pandas.util import testing as tm
-
 from pandas._libs.tslibs import parsing
 
 
@@ -68,7 +66,7 @@ def test_parsers_monthfreq(self):
 
 class TestGuessDatetimeFormat(object):
 
-    @pytest.mark.xfail(reason="GH18141 - dateutil > 2.6.1 broken")
+    @is_dateutil_le_261
     @pytest.mark.parametrize(
         "string, format",
         [
@@ -86,7 +84,20 @@ def test_guess_datetime_format_with_parseable_formats(
         result = parsing._guess_datetime_format(string)
         assert result == format
 
-    @pytest.mark.xfail(reason="GH18141 - dateutil > 2.6.1 broken")
+    @is_dateutil_gt_261
+    @pytest.mark.parametrize(
+        "string",
+        ['20111230', '2011-12-30', '30-12-2011',
+         '2011-12-30 00:00:00', '2011-12-30T00:00:00',
+         '2011-12-30 00:00:00.000000'])
+    def test_guess_datetime_format_with_parseable_formats_gt_261(
+            self, string):
+        tm._skip_if_not_us_locale()
+
+        result = parsing._guess_datetime_format(string)
+        assert result is None
+
+    @is_dateutil_le_261
     @pytest.mark.parametrize(
         "dayfirst, expected",
         [
@@ -98,7 +109,16 @@ def test_guess_datetime_format_with_dayfirst(self, dayfirst, expected):
             ambiguous_string, dayfirst=dayfirst)
         assert result == expected
 
-    @pytest.mark.xfail(reason="GH18141 - dateutil > 2.6.1 broken")
+    @is_dateutil_gt_261
+    @pytest.mark.parametrize(
+        "dayfirst", [True, False])
+    def test_guess_datetime_format_with_dayfirst_gt_261(self, dayfirst):
+        ambiguous_string = '01/01/2011'
+        result = parsing._guess_datetime_format(
+            ambiguous_string, dayfirst=dayfirst)
+        assert result is None
+
+    @is_dateutil_le_261
     @pytest.mark.parametrize(
         "string, format",
         [
@@ -114,6 +134,22 @@ def test_guess_datetime_format_with_locale_specific_formats(
         result = parsing._guess_datetime_format(string)
         assert result == format
 
+    @is_dateutil_gt_261
+    @pytest.mark.parametrize(
+        "string",
+        [
+            '30/Dec/2011',
+            '30/December/2011',
+            '30/Dec/2011 00:00:00'])
+    def test_guess_datetime_format_with_locale_specific_formats_gt_261(
+            self, string):
+        # The month names will vary depending on the locale, in which
+        # case these wont be parsed properly (dateutil can't parse them)
+        tm._skip_if_has_locale()
+
+        result = parsing._guess_datetime_format(string)
+        assert result is None
+
     def test_guess_datetime_format_invalid_inputs(self):
         # A datetime string must include a year, month and a day for it
         # to be guessable, in addition to being a string that looks like
@@ -132,7 +168,7 @@ def test_guess_datetime_format_invalid_inputs(self):
         for invalid_dt in invalid_dts:
             assert parsing._guess_datetime_format(invalid_dt) is None
 
-    @pytest.mark.xfail(reason="GH18141 - dateutil > 2.6.1 broken")
+    @is_dateutil_le_261
     @pytest.mark.parametrize(
         "string, format",
         [
@@ -147,6 +183,21 @@ def test_guess_datetime_format_nopadding(self, string, format):
         result = parsing._guess_datetime_format(string)
         assert result == format
 
+    @is_dateutil_gt_261
+    @pytest.mark.parametrize(
+        "string",
+        [
+            '2011-1-1',
+            '30-1-2011',
+            '1/1/2011',
+            '2011-1-1 00:00:00',
+            '2011-1-1 0:0:0',
+            '2011-1-3T00:00:0'])
+    def test_guess_datetime_format_nopadding_gt_261(self, string):
+        # GH 11142
+        result = parsing._guess_datetime_format(string)
+        assert result is None
+
 
 class TestArrayToDatetime(object):
     def test_try_parse_dates(self):