Issue 54781

Summary: Writer: yet another problem with merging of text attributes
Product: Writer Reporter: thomas.lange
Component: codeAssignee: AOO issues mailing list <issues>
Status: CONFIRMED --- QA Contact:
Severity: Trivial    
Priority: P3 CC: issues
Version: current   
Target Milestone: ---   
Hardware: All   
OS: All   
Issue Type: DEFECT Latest Confirmation in: ---
Developer Difficulty: ---
Attachments:
Description Flags
Sample bug doc
none
Yet another document with attribute problems none

Description thomas.lange 2005-09-19 09:17:04 UTC
There is another problem when merging text attributes that from the viewpoint of
the user leads to some arbitrary behaviour.
Comment 1 thomas.lange 2005-09-19 09:17:52 UTC
Created attachment 29657 [details]
Sample bug doc
Comment 2 thomas.lange 2005-09-19 09:20:34 UTC
If you open the attached bugdoc and start simplified -> traditional conversion
the re is achance that the "4" in the simplified line remains with language
simplified and/or that the "." at the end of the same line is traditional, the
last character before it as well, but if you select both of them the language is
not known anymore.

Of course the same problems may occur for other attributes as well.
Comment 3 frank.meies 2005-09-19 10:23:12 UTC
FME: Another way to reproduce the bug is this:

1) Type "Hello World"
2) Select all, select font A
3) Select "Hello", select font B
4) Select the blank between "Hello" and "World" and choose Formatting -> Default
Formatting

We now have this situation: "Hello" has two attributes: font A and font B. It is
not predictible which of these is applied.

Solution: In SwTxtNode::RstAttr(), not only the second part of the attribute to
split should be checked if it should be ignored (using Forget()), but also the
first part of the split attribute should be checked.
Comment 4 frank.meies 2005-09-20 10:39:56 UTC
FME: Set target to 'OOo later'.
Comment 5 thomas.lange 2005-09-29 12:53:46 UTC
Created attachment 29993 [details]
Yet another document with attribute problems
Comment 6 thomas.lange 2005-09-29 12:55:48 UTC
I added another bugdocument (Karl1-computer.odt) if you open this one and do a
traditional->simplified conversion with "translate common terms" enabled
sometimes the first two characters become italic. Something that should not happen.