BUG: read_excel(engine='openpyxl') with "unsized" XLSX issue

[kuraga]

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

Problem description

Consider code:

import pandas as pd
pd.read_excel('test.xlsx', index_col=[0, 1], engine='openpyxl')

with some special test.xlsx (see below).

According to openpyxl issue #1483 and openpyxl documentation, user (i.e. Pandas) have to call sheet.calculate_dimension(force=True) if any workbook's worksheet is "unsized".

(What is "worksheet is unsized"? It means: the worksheet doesn't have DIMENSION_TAG, see WorkSheetParser.parse_dimensions(). When it checked? On ReadOnlyWorksheet object construction.)

But Pandas doesn't do it.

So, if the worksheet is "unsized", on reading it by read_excel(engine='openpyxl') we try to get sheet data. We iterate rows falling down into openpyxl: 1 -> 2 -> 3 -> 4 (with max_col == None) -> 5 -> 6 (it returns self._max_column == None) -> 7 -> 8 (with max_col == None) -> 9 -> 10 (with max_col == None). And in these lines:

max_col = max_col or  row[-1]['column']
row_width = max_col + 1 - min_col

we have: row's cells number (row_width) is individual for each row (row).

Now consider an unsized Excel table (I have a real file but it's private; after saving in LibreOffice it's become sized):

We get sheet data:

[ [`A1`],
  [ 'B1', 'B2' ] ]

But index_col=[0, 1] suppose to have >=2 cols in each row:

Traceback (most recent call last):
  File "test.py", line 6, in <module>
    pd.read_excel('test.xlsx', index_col=[0, 1], engine='openpyxl')
  File "/home/sasha/miniconda3/lib/python3.7/site-packages/pandas/io/excel/_base.py", line 334, in read_excel
    **kwds,
  File "/home/sasha/miniconda3/lib/python3.7/site-packages/pandas/io/excel/_base.py", line 888, in parse
    **kwds,
  File "/home/sasha/miniconda3/lib/python3.7/site-packages/pandas/io/excel/_base.py", line 480, in parse
    last = data[offset][col]
IndexError: list index out of range

IMHO, patch could be:

diff --git a/pandas/io/excel/_openpyxl.py b/pandas/io/excel/_openpyxl.py
index c4327316d..3efbf4abc 100644
--- a/pandas/io/excel/_openpyxl.py
+++ b/pandas/io/excel/_openpyxl.py
@@ -536,6 +536,9 @@ class _OpenpyxlReader(_BaseExcelReader):
 
     def get_sheet_data(self, sheet, convert_float: bool) -> List[List[Scalar]]:
         data: List[List[Scalar]] = []
+
+        sheet.calculate_dimension(force=True)
+
         for row in sheet.rows:
             data.append([self._convert_cell(cell, convert_float) for cell in row])

(But it may provide some drawbacks).
With this patch, sheet data would be:

[ [`A1`, ''],
  [ 'B1', 'B2' ] ]

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.7.7.final.0
python-bits : 64
OS : Linux
OS-release : 5.4.42-calculate
machine : x86_64
processor : Intel(R) Core(TM) i5-7200U CPU @ 2.50GHz
byteorder : little
LC_ALL : None
LANG : ru_RU.utf8
LOCALE : ru_RU.UTF-8

pandas : 1.0.3
numpy : 1.18.1
pytz : 2020.1
dateutil : 2.8.1
pip : 20.0.2
setuptools : 46.4.0.post20200518
Cython : 0.29.17
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : 1.2.8
lxml.etree : 4.5.0
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 2.11.2
IPython : 7.13.0
pandas_datareader: None
bs4 : None
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : 4.5.0
matplotlib : 3.1.3
numexpr : None
odfpy : None
openpyxl : 3.0.3
pandas_gbq : None
pyarrow : None
pytables : None
pytest : None
pyxlsb : None
s3fs : None
scipy : 1.4.1
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : 1.2.0
xlwt : None
xlsxwriter : 1.2.8
numba : None

Thanks!

[jbrockmendel]

According to openpyxl issue #1483 and openpyxl documentation, user (i.e. Pandas) have to call sheet.calculate_dimension(force=True) if any workbook's worksheet is "unsized".

Want to make a PR to do this?

[theoniko]

take

[asishm]

@kuraga would you be able to check if #39486 fixes this?

[rhshadrach]

@kuraga - My expectation is yes, but without a sheet to test it on there is nothing we can do. I'm going to mark this accordingly but if we get no response, my thought is to close.

[kuraga]

Sorry don't have sources now.

[rhshadrach]

@kuraga - not a problem, thanks for the response.