Merge pull request #13 from PLPeeters/bugfix/column-to-number

Fix `column_to_number` being incorrect for files with more than 26 columns

with big thanks!

Author: chfw

.github/workflows/tests.yml

@@ -29,7 +29,7 @@ jobs:
       - name: test
         run: |
           pip freeze
-          nosetests --verbosity=3 --with-coverage --cover-package pyexcel_xlsxr --cover-package tests tests --with-doctest --doctest-extension=.rst README.rst  pyexcel_xlsxr
+          coverage run -m --source=pyexcel_xlsxr pytest && coverage report --show-missing
       - name: Upload coverage
         uses: codecov/codecov-action@v1
         with:

CONTRIBUTORS.rst

@@ -6,3 +6,4 @@
 In alphabetical order:
 
 * `Mark Skelton <https://github.com/mtskelton>`_
+* `Pierre-Louis Peeters <https://github.com/PLPeeters>`_

changelog.yml

@@ -2,7 +2,12 @@ name: pyexcel-xlsxr
 organisation: pyexcel
 releases:
 - changes:
-  - action: Updated
+- action: Fixed
+    details:
+    - 'Fix reading of files with more than 26 columns'
+  date: 30.10.2025
+  version: 0.6.2
+- action: Updated
     details:
     - '#9: Potential fix for incorrect reading of data with empty cells when used with pyexcel '
   date: 11.11.2024

lint.sh

@@ -1,2 +1,2 @@
 pip install flake8
-flake8 --exclude=.moban.d,docs,setup.py   --builtins=unicode,xrange,long .  && python setup.py checkdocs
+flake8 --exclude=.moban.d,docs,setup.py,.venv   --builtins=unicode,xrange,long .  && python setup.py checkdocs

pyexcel_xlsxr/__init__.py

@@ -1,10 +1,11 @@
 """
-    pyexcel_xlsxr
-    ~~~~~~~~~~~~~~~~~~~
-    The lower level xlsx file format handler using lxml
-    :copyright: (c) 2015-2020 by Onni Software Ltd & its contributors
-    :license: New BSD License
+pyexcel_xlsxr
+~~~~~~~~~~~~~~~~~~~
+The lower level xlsx file format handler using lxml
+:copyright: (c) 2015-2020 by Onni Software Ltd & its contributors
+:license: New BSD License
 """
+
 from pyexcel_io.io import get_data as read_data
 from pyexcel_io.io import isstream
 from pyexcel_io.plugins import IOPluginInfoChainV2

pyexcel_xlsxr/_version.py

@@ -1,2 +1,2 @@
-__version__ = '0.6.1'
-__author__ = 'C.W.'
+__version__ = "0.6.1"
+__author__ = "C.W."

pyexcel_xlsxr/messy_xlsx.py

@@ -2,6 +2,7 @@
 import re
 import zipfile
 from datetime import time, datetime, timedelta
+from functools import cache
 
 from lxml import etree
 from pyexcel_io._compact import OrderedDict
@@ -182,12 +183,16 @@ def __repr__(self):
         return str(self.value)
 
 
+@cache
 def column_to_number(column):
-    column = re.sub("[^A-Z]", "", column)
-    cl = len(column) - 1
-    return sum(
-        [(ord(c.upper()) - 64) + (26 * (cl - i)) for i, c in enumerate(column)]
-    )
+    column = re.sub(r"[^A-Z]", "", column.upper())
+
+    result = 0
+
+    for index, c in enumerate(column):
+        result = result * 26 + (ord(c) - ord("A") + 1)
+
+    return result
 
 
 def parse_row(row_xml_string, book):
@@ -338,10 +343,10 @@ def parse_book_properties(book_content):
     )
     namespaces = {"r": ns}
 
-    xlsx_header = u"<wrapper {0}>".format(
+    xlsx_header = "<wrapper {0}>".format(
         " ".join('xmlns:{0}="{1}"'.format(k, v) for k, v in namespaces.items())
     ).encode("utf-8")
-    xlsx_footer = u"</wrapper>".encode("utf-8")
+    xlsx_footer = "</wrapper>".encode("utf-8")
     sheets = SHEET_FMT_MATCHER.findall(book_content)
     for sheet in sheets:
         block = xlsx_header + sheet + xlsx_footer

test.bat

@@ -1,2 +1,2 @@
 pip freeze
-nosetests --with-coverage --cover-package pyexcel_xlsxr --cover-package tests tests --with-doctest --doctest-extension=.rst README.rst  pyexcel_xlsxr
+coverage run -m --source=pyexcel_xlsxr pytest && coverage report --show-missing

test.sh

@@ -1,3 +1,3 @@
 #/bin/bash
 pip freeze
-nosetests --with-coverage --cover-package pyexcel_xlsxr --cover-package tests tests --with-doctest --doctest-extension=.rst README.rst  pyexcel_xlsxr
+coverage run -m --source=pyexcel_xlsxr pytest && coverage report --show-missing

tests/base.py

@@ -1,10 +1,5 @@
-import os  # noqa
-import datetime  # noqa
-
 import pyexcel
 
-from nose.tools import eq_, raises  # noqa
-
 
 def create_sample_file1(file):
     data = ["a", "b", "c", "d", "e", "f", "g", "h", "i", "j", 1.1, 1]