#878330 invalid heap memory read and garbage output when parsing XLS files with international characters

Package:
catdoc
Source:
catdoc
Description:
text extractor for MS-Office files
Submitter:
Robert Zavalczki
Date:
2025-06-12 09:31:01 UTC
Severity:
normal
Tags:
#878330#5
Date:
2017-10-12 21:37:31 UTC
From:
To:
This issue happens when running 'xls2csv' on XLS files containing international characters. The issue can be reproduced with the attached 'unicode_MS.xls' file.

Details: When investigating more I noticed that sometimes contents from the xls2csv process heap memory was dumped to the output.
Note: Applying the attached patch to the source code in version 0.95.4 seems to fix the issue.
#878330#10
Date:
2025-06-12 08:49:54 UTC
From:
To:
On Fri, 13 Oct 2017 00:37:31 +0300 Robert Zavalczki  <robert.zavalczki@gmail.com> wrote:

 > This issue happens when running 'xls2csv' on XLS files containing
international characters. The issue can be reproduced with the attached
'unicode_MS.xls' file.
 > ...
 > Note: Applying the attached patch to the source code in version
0.95.4 seems to fix the issue.

FYI I incorporated your test file and patch in my fork
https://github.com/skierpage/catdoc, and verified that the patch fixes
the bug with the test file.