DEP: bump min version of openpyxl to 3.0.0 #39603 (#39702)

Author: fangchenli

ci/deps/azure-37-locale_slow.yaml

@@ -18,7 +18,7 @@ dependencies:
   - lxml
   - matplotlib=3.0.0
   - numpy=1.16.*
-  - openpyxl=2.6.0
+  - openpyxl=3.0.0
   - python-dateutil
   - python-blosc
   - pytz=2017.3

ci/deps/azure-37-minimum_versions.yaml

@@ -19,7 +19,7 @@ dependencies:
   - numba=0.46.0
   - numexpr=2.6.8
   - numpy=1.16.5
-  - openpyxl=2.6.0
+  - openpyxl=3.0.0
   - pytables=3.5.1
   - python-dateutil=2.7.3
   - pytz=2017.3

doc/source/getting_started/install.rst

@@ -274,7 +274,7 @@ html5lib                  1.0.1              HTML parser for read_html (see :ref
 lxml                      4.3.0              HTML parser for read_html (see :ref:`note <optional_html>`)
 matplotlib                2.2.3              Visualization
 numba                     0.46.0             Alternative execution engine for rolling operations
-openpyxl                  2.6.0              Reading / writing for xlsx files
+openpyxl                  3.0.0              Reading / writing for xlsx files
 pandas-gbq                0.12.0             Google Big Query access
 psycopg2                  2.7                PostgreSQL engine for sqlalchemy
 pyarrow                   0.15.0             Parquet, ORC, and feather reading / writing

doc/source/whatsnew/v1.3.0.rst

@@ -186,7 +186,7 @@ Optional libraries below the lowest tested version may still work, but are not c
 +-----------------+-----------------+---------+
 | numba           | 0.46.0          |         |
 +-----------------+-----------------+---------+
-| openpyxl        | 2.6.0           |         |
+| openpyxl        | 3.0.0           |    X    |
 +-----------------+-----------------+---------+
 | pyarrow         | 0.15.0          |         |
 +-----------------+-----------------+---------+

pandas/compat/_optional.py

@@ -17,7 +17,7 @@
     "matplotlib": "2.2.3",
     "numexpr": "2.6.8",
     "odfpy": "1.3.0",
-    "openpyxl": "2.6.0",
+    "openpyxl": "3.0.0",
     "pandas_gbq": "0.12.0",
     "pyarrow": "0.15.0",
     "pytest": "5.0.1",

pandas/io/excel/_openpyxl.py

@@ -1,13 +1,12 @@
 from __future__ import annotations
 
-from distutils.version import LooseVersion
 import mmap
 from typing import TYPE_CHECKING, Dict, List, Optional
 
 import numpy as np
 
 from pandas._typing import FilePathOrBuffer, Scalar, StorageOptions
-from pandas.compat._optional import get_version, import_optional_dependency
+from pandas.compat._optional import import_optional_dependency
 
 from pandas.io.excel._base import BaseExcelReader, ExcelWriter
 from pandas.io.excel._util import validate_freeze_panes
@@ -531,14 +530,8 @@ def _convert_cell(self, cell, convert_float: bool) -> Scalar:
         return cell.value
 
     def get_sheet_data(self, sheet, convert_float: bool) -> List[List[Scalar]]:
-        # GH 39001
-        # Reading of excel file depends on dimension data being correct but
-        # writers sometimes omit or get it wrong
-        import openpyxl
 
-        version = LooseVersion(get_version(openpyxl))
-
-        if version >= "3.0.0" and self.book.read_only:
+        if self.book.read_only:
             sheet.reset_dimensions()
 
         data: List[List[Scalar]] = []
@@ -552,7 +545,7 @@ def get_sheet_data(self, sheet, convert_float: bool) -> List[List[Scalar]]:
         # Trim trailing empty rows
         data = data[: last_row_with_data + 1]
 
-        if version >= "3.0.0" and self.book.read_only and len(data) > 0:
+        if self.book.read_only and len(data) > 0:
             # With dimension reset, openpyxl no longer pads rows
             max_width = max(len(data_row) for data_row in data)
             if min(len(data_row) for data_row in data) < max_width:

pandas/tests/io/excel/test_openpyxl.py

@@ -1,11 +1,8 @@
-from distutils.version import LooseVersion
 from pathlib import Path
 
 import numpy as np
 import pytest
 
-from pandas.compat._optional import get_version
-
 import pandas as pd
 from pandas import DataFrame
 import pandas._testing as tm
@@ -157,10 +154,6 @@ def test_read_with_bad_dimension(
     datapath, ext, header, expected_data, filename, read_only, request
 ):
     # GH 38956, 39001 - no/incorrect dimension information
-    version = LooseVersion(get_version(openpyxl))
-    if (read_only or read_only is None) and version < "3.0.0":
-        msg = "openpyxl read-only sheet is incorrect when dimension data is wrong"
-        request.node.add_marker(pytest.mark.xfail(reason=msg))
     path = datapath("io", "data", "excel", f"{filename}{ext}")
     if read_only is None:
         result = pd.read_excel(path, header=header)
@@ -195,10 +188,6 @@ def test_append_mode_file(ext):
 @pytest.mark.parametrize("read_only", [True, False, None])
 def test_read_with_empty_trailing_rows(datapath, ext, read_only, request):
     # GH 39181
-    version = LooseVersion(get_version(openpyxl))
-    if (read_only or read_only is None) and version < "3.0.0":
-        msg = "openpyxl read-only sheet is incorrect when dimension data is wrong"
-        request.node.add_marker(pytest.mark.xfail(reason=msg))
     path = datapath("io", "data", "excel", f"empty_trailing_rows{ext}")
     if read_only is None:
         result = pd.read_excel(path)