Bug 85648 - FILEOPEN HTML inside .xls saved using AOpenOffice 4.1.1 opens in Libre Writer 4.3.2
Summary: FILEOPEN HTML inside .xls saved using AOpenOffice 4.1.1 opens in Libre Writer...
Status: RESOLVED WORKSFORME
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Calc (show other bugs)
Version:
(earliest affected)
4.2.7.2 release
Hardware: x86-64 (AMD64) Windows (All)
: medium normal
Assignee: Not Assigned
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-10-30 12:46 UTC by Mateo
Modified: 2017-07-12 20:31 UTC (History)
3 users (show)

See Also:
Crash report or crash signature:


Attachments
html saved as xls (95.14 KB, text/html)
2014-10-30 12:46 UTC, Mateo
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Mateo 2014-10-30 12:46:35 UTC
Created attachment 108688 [details]
html saved as xls

I am trying to move from Open Office 4.1.1 to the Libre 4.3.2. The department I tried this with has a process where they save information from a website into OpenOffice CALC and then save it as .xls.  These files can NOT be opened using Libre CALC (4.3.2) with the .xls extension. If I change the extension to .html it of course opens by default in a browser just fine. If I right-click and open the .html file with Libre CALC it opens the Import Options dialog box and I just hit OK and it opens fine inside Libre Calc. (Font default as Liberation Serif but thats another issue). The bug is that Open Office opens these files just fine as they are (.xls) and Libre Office (4.3.2) does not. It appears that there were bugs with this same feature in previous versions but they said the newer code would have the patches built in....
Comment 1 Maxim Monastirsky 2014-10-30 13:32:11 UTC
Hi msaum,

Thanks for the test file, I'll investigate soon.
Comment 2 Maxim Monastirsky 2014-10-30 20:11:46 UTC
@David: Hi, Can I cherry-pick [1] to 4-3? In that case I guess it worth also adding [2], to make sure we don't create a regression for --convert-to?

[1] http://cgit.freedesktop.org/libreoffice/core/commit/?id=86c6f18c2766aad43d6e3bfcf3530e40440ebca7
[2] http://cgit.freedesktop.org/libreoffice/core/commit/?id=24d20bce789063f6dba2df1c5d2c5a8948d24370
Comment 3 Julio Gázquez 2015-03-06 13:20:02 UTC
@Maxim those patches fix the issue?

I can provide some aditional details:

It's a regression, it worked in LO in versions up to 3.6.x at least, in 4.2.x the bug is already there. I'm not sure if, as Mateo said, the bug was introduced while fixing other issues.

It's interesting that LO is being being able to detect and open properly this kind of files, just it's not detecting *some* of those files.

The bug is fired if the file contains certain special characters (or bytes), in 8 bit encodings.

In Mateo's sample file, it's Windows-1252 "trademark" character (byte of value 0x99). Removing the ofending character the file opens fine.

In my particular case (also a 3rd party app mimicking Excel generated HTML files), the file is encoded as as iso8859-1, failing with "masculine ordinal" character (byte of value 0xba)

I also found that opening the file in a text editor, changing declared encoding to UTF-8 and saving as UTF-8 makes LibreOffice open the file properly, as a Calc spreadsheet.
Comment 4 Mateo 2015-04-22 17:21:00 UTC
Will this be fixed in 4.3.x.x because it doesn't work in version 4.3.6.2 either? It is fixed in 4.4.x.x, however that version introduces a problem with the horizontal bar being "missing" when viewed by a MS Excel user.(Bug 89058) I guess I am stuck waiting for 4.5? OR whenever OpenOffice gets their next version out OR MS Office.
Comment 5 QA Administrators 2016-09-20 09:32:02 UTC Comment hidden (obsolete)
Comment 6 Mateo 2016-10-25 13:08:27 UTC
Close this issue. It seems to be fine in LO 5.2.2