Merge pull request pandas-dev#8 from dimastbk/issue-50395

Author: kostyafarber

ci/deps/actions-311.yaml

@@ -55,5 +55,5 @@ dependencies:
   - zstandard>=0.15.2
 
   - pip:
-    - python-calamine>=0.1.0
+    - python-calamine>=0.1.1
     - tzdata>=2022.1

ci/deps/actions-38-minimum_versions.yaml

@@ -59,5 +59,5 @@ dependencies:
 
   - pip:
     - pyqt5==5.15.1
-    - python-calamine==0.1.0
+    - python-calamine==0.1.1
     - tzdata==2022.1

doc/source/getting_started/install.rst

@@ -345,7 +345,7 @@ xlrd                      2.0.1              excel           Reading Excel
 xlsxwriter                1.4.3              excel           Writing Excel
 openpyxl                  3.0.7              excel           Reading / writing for xlsx files
 pyxlsb                    1.0.8              excel           Reading for xlsb files
-python-calamine           0.1.0              excel           Reading for xls/xlsx/xlsb/ods files
+python-calamine           0.1.1              excel           Reading for xls/xlsx/xlsb/ods files
 ========================= ================== =============== =============================================================
 
 HTML

doc/source/user_guide/io.rst

@@ -3421,7 +3421,8 @@ The :func:`~pandas.read_excel` method can read Excel 2007+ (``.xlsx``) files
 using the ``openpyxl`` Python module. Excel 2003 (``.xls``) files
 can be read using ``xlrd``. Binary Excel (``.xlsb``)
 files can be read using ``pyxlsb``. Also, all this formats can be read using ``python-calamine``,
-but this library has sime limitation, for example, can't detect date in most formats.
+but this library has some limitation and different behavior from other libraries,
+for example, can't detect date in some formats (xls and xlsb).
 The :meth:`~DataFrame.to_excel` instance method is used for
 saving a ``DataFrame`` to Excel.  Generally the semantics are
 similar to working with :ref:`csv<io.read_csv_table>` data.

pandas/compat/_optional.py

@@ -37,7 +37,7 @@
     "pyarrow": "7.0.0",
     "pyreadstat": "1.1.2",
     "pytest": "7.0.0",
-    "python-calamine": "0.1.0",
+    "python-calamine": "0.1.1",
     "pyxlsb": "1.0.8",
     "s3fs": "2021.08.0",
     "scipy": "1.7.1",

pandas/tests/io/excel/test_readers.py

@@ -174,10 +174,10 @@ def test_usecols_list(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -204,10 +204,10 @@ def test_usecols_str(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -268,10 +268,10 @@ def test_usecols_diff_positional_int_columns_order(
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -296,10 +296,10 @@ def test_read_excel_without_slicing(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -314,10 +314,10 @@ def test_usecols_excel_range_str(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -400,11 +400,9 @@ def test_excel_cell_error_na(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".ods"}:
+        if engine == "calamine" and read_ext == ".ods":
             request.node.add_marker(
-                pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
-                )
+                pytest.mark.xfail(reason="Calamine returns 0 instead of NaN in ods")
             )
 
         parsed = pd.read_excel("test3" + read_ext, sheet_name="Sheet1")
@@ -418,10 +416,10 @@ def test_excel_table(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -445,15 +443,15 @@ def test_reader_special_dtypes(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
-        if engine == "calamine":
+        if engine == "calamine" and read_ext != ".ods":
             request.node.add_marker(
-                pytest.mark.xfail(reason="Calamine can't parse this datetime format")
+                pytest.mark.xfail(reason="Maybe not supported by calamine")
             )
 
         expected = DataFrame.from_dict(
@@ -715,13 +713,6 @@ def test_dtype_mangle_dup_cols(self, read_ext, dtypes, exp_value):
 
     def test_reader_spaces(self, request, engine, read_ext):
         # see gh-32207
-
-        # https://github.com/tafia/calamine/pull/289
-        if engine == "calamine" and read_ext == ".ods":
-            request.node.add_marker(
-                pytest.mark.xfail(reason="Calamine doesn't respect spaces in ods")
-            )
-
         basename = "test_spaces"
 
         actual = pd.read_excel(basename + read_ext)
@@ -826,7 +817,7 @@ def test_date_conversion_overflow(self, request, engine, read_ext):
             columns=["DateColWithBigInt", "StringCol"],
         )
 
-        if engine == "calamine":
+        if engine == "calamine" and read_ext != ".ods":
             request.node.add_marker(
                 pytest.mark.xfail(reason="Maybe not supported by calamine")
             )
@@ -852,10 +843,10 @@ def test_sheet_name(self, request, read_ext, engine, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -1027,10 +1018,10 @@ def test_reader_seconds(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
         if engine == "calamine":
@@ -1073,12 +1064,16 @@ def test_read_excel_multiindex(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
+        if engine == "calamine" and read_ext == ".ods":
+            request.node.add_marker(
+                pytest.mark.xfail(reason="Last test fails in calamine")
+            )
 
         mi = MultiIndex.from_product([["foo", "bar"], ["a", "b"]])
         mi_file = "testmultiindex" + read_ext
@@ -1175,10 +1170,10 @@ def test_read_excel_multiindex_blank_after_name(
                     f"not supported by {engine} (GH4679)"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -1310,10 +1305,10 @@ def test_read_excel_skiprows(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -1371,10 +1366,10 @@ def test_read_excel_skiprows_callable_not_in(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -1651,10 +1646,10 @@ def test_excel_table_sheet_by_index(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -1686,10 +1681,10 @@ def test_sheet_name(self, request, engine, read_ext, df_ref):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 
@@ -1784,10 +1779,10 @@ def test_read_datetime_multiindex(self, request, engine, read_ext):
                     reason="Sheets containing datetimes not supported by pyxlsb"
                 )
             )
-        if engine == "calamine" and read_ext in {".xls", ".xlsb", ".ods"}:
+        if engine == "calamine" and read_ext in {".xls", ".xlsb"}:
             request.node.add_marker(
                 pytest.mark.xfail(
-                    reason="Calamine support parsing datetime only in xlsx"
+                    reason="Calamine support parsing datetime only in xlsx/ods"
                 )
             )
 

pyproject.toml

@@ -62,7 +62,7 @@ computation = ['scipy>=1.7.1', 'xarray>=0.21.0']
 fss = ['fsspec>=2021.07.0']
 aws = ['s3fs>=2021.08.0']
 gcp = ['gcsfs>=2021.07.0', 'pandas-gbq>=0.15.0']
-excel = ['odfpy>=1.4.1', 'openpyxl>=3.0.7', 'python-calamine>=0.1.0', 'pyxlsb>=1.0.8', 'xlrd>=2.0.1', 'xlsxwriter>=1.4.3']
+excel = ['odfpy>=1.4.1', 'openpyxl>=3.0.7', 'python-calamine>=0.1.1', 'pyxlsb>=1.0.8', 'xlrd>=2.0.1', 'xlsxwriter>=1.4.3']
 parquet = ['pyarrow>=7.0.0']
 feather = ['pyarrow>=7.0.0']
 hdf5 = [# blosc only available on conda (https://github.com/Blosc/python-blosc/issues/297)
@@ -104,7 +104,7 @@ all = ['beautifulsoup4>=4.9.3',
        'pytest>=7.0.0',
        'pytest-xdist>=2.2.0',
        'pytest-asyncio>=0.17.0',
-       'python-calamine>=0.1.0',
+       'python-calamine>=0.1.1',
        'python-snappy>=0.6.0',
        'pyxlsb>=1.0.8',
        'qtpy>=2.2.0',

scripts/tests/data/deps_expected_random.yaml

@@ -57,4 +57,4 @@ dependencies:
   - zstandard>=0.15.2
 
   - pip:
-    - python-calamine>=0.1.0
+    - python-calamine>=0.1.1

scripts/tests/data/deps_minimum.toml

@@ -62,7 +62,7 @@ computation = ['scipy>=1.7.1', 'xarray>=0.21.0']
 fss = ['fsspec>=2021.07.0']
 aws = ['s3fs>=2021.08.0']
 gcp = ['gcsfs>=2021.07.0', 'pandas-gbq>=0.15.0']
-excel = ['odfpy>=1.4.1', 'openpyxl>=3.0.7', 'python-calamine>=0.1.0', 'pyxlsb>=1.0.8', 'xlrd>=2.0.1', 'xlsxwriter>=1.4.3']
+excel = ['odfpy>=1.4.1', 'openpyxl>=3.0.7', 'python-calamine>=0.1.1', 'pyxlsb>=1.0.8', 'xlrd>=2.0.1', 'xlsxwriter>=1.4.3']
 parquet = ['pyarrow>=7.0.0']
 feather = ['pyarrow>=7.0.0']
 hdf5 = [# blosc only available on conda (https://github.com/Blosc/python-blosc/issues/297)
@@ -104,7 +104,7 @@ all = ['beautifulsoup4>=5.9.3',
        'pytest>=7.0.0',
        'pytest-xdist>=2.2.0',
        'pytest-asyncio>=0.17.0',
-       'python-calamine>=0.1.0',
+       'python-calamine>=0.1.1',
        'python-snappy>=0.6.0',
        'pyxlsb>=1.0.8',
        'qtpy>=2.2.0',

scripts/tests/data/deps_unmodified_random.yaml

@@ -57,4 +57,4 @@ dependencies:
   - zstandard>=0.15.2
 
   - pip:
-    - python-calamine>=0.1.0
+    - python-calamine>=0.1.1