Issue 58366 - rounding issues with text positions in PDF export
Summary: rounding issues with text positions in PDF export
Status: CLOSED FIXED
Alias: None
Product: gsl
Classification: Code
Component: code (show other issues)
Version: 680m139
Hardware: PC All
: P3 Trivial (vote)
Target Milestone: OOo 2.0.1
Assignee: ulf.stroehler
QA Contact: issues@gsl
URL:
Keywords: oooqa
Depends on:
Blocks:
 
Reported: 2005-11-23 19:51 UTC by novomesk
Modified: 2005-11-28 08:16 UTC (History)
3 users (show)

See Also:
Issue Type: DEFECT
Latest Confirmation in: ---
Developer Difficulty: ---


Attachments
Here you can find the mentioned 2100 with 00 in superscript (7.30 KB, application/vnd.oasis.opendocument.text)
2005-11-23 19:52 UTC, novomesk
no flags Details
Here you can see the problem (22.97 KB, application/pdf)
2005-11-23 19:54 UTC, novomesk
no flags Details
a screenshot the problem (54.78 KB, image/png)
2005-11-24 11:49 UTC, novomesk
no flags Details
hotfix to solve this problem (433 bytes, patch)
2005-11-24 17:03 UTC, philipp.lohmann
no flags Details | Diff

Note You need to log in before you can comment on or make changes to this issue.
Description novomesk 2005-11-23 19:51:11 UTC
I am attaching a short document with a single sentence.
The word 2100 has the two last zeros in superscript. The two zeros have wrong
position in the exported PDF.

I have this problem with OO-dev 139 on Windows
but in OO 2.0.0 on Linux everything is OK.
Comment 1 novomesk 2005-11-23 19:52:33 UTC
Created attachment 31747 [details]
Here you can find the mentioned 2100 with 00 in superscript
Comment 2 novomesk 2005-11-23 19:54:01 UTC
Created attachment 31748 [details]
Here you can see the problem
Comment 3 lars 2005-11-23 21:50:03 UTC
when I export the odt as pdf in OOo-dev 2.0.139 the superscript 00 is at the 
correct position.
Comment 4 novomesk 2005-11-24 06:43:37 UTC
In my exported PDF, the first zero in superscript is intersecting the previous
character '1' 
Older version are putting some space between those two characters, so they are
not intersecting.

The behavior I described is correct?
Please can you post the PDF exported by your 139, I would like to compare the
results.
Comment 5 novomesk 2005-11-24 11:49:20 UTC
Created attachment 31765 [details]
a screenshot the problem
Comment 6 hdu@apache.org 2005-11-24 15:05:50 UTC
I can reproduce it with a 680m142, but not with a 680m135...
Comment 7 stephan_schaefer 2005-11-24 15:10:29 UTC
ssa->hdu: please take over.
Comment 8 hdu@apache.org 2005-11-24 16:22:05 UTC
Isolated it to the changes for pdfwriter_impl.*xx in CWS vcl42. Looks like the
old "accummulated advance width" rounding error.

HDU->PL: please take over
Comment 9 philipp.lohmann 2005-11-24 17:03:11 UTC
Created attachment 31778 [details]
hotfix to solve this problem
Comment 10 philipp.lohmann 2005-11-24 17:15:19 UTC
Superscript is only the tip of the iceberg; e.g. justified text is totally
botched. This broke with my fix for issue 49748 which makes rounding for lines
better but for text equally worse. The best solution for 2.0.1 would be to back
out the fix for 49748 since it is a much smaller problem and fix that one for 2.0.2.
Comment 11 philipp.lohmann 2005-11-25 10:24:42 UTC
fixed in CWS src680m142masterfix
Comment 12 philipp.lohmann 2005-11-25 10:25:34 UTC
pl->us: please verify in new m142 master

re-open issue and reassign to us@openoffice.org
Comment 13 philipp.lohmann 2005-11-25 10:25:38 UTC
reassign to us@openoffice.org
Comment 14 philipp.lohmann 2005-11-25 10:25:42 UTC
reset resolution to FIXED
Comment 15 philipp.lohmann 2005-11-25 10:26:10 UTC
reassign to us@openoffice.org
Comment 16 philipp.lohmann 2005-11-25 10:26:14 UTC
reset resolution to FIXED
Comment 17 ulf.stroehler 2005-11-28 08:15:49 UTC
Verified in m142 latest.
Comment 18 ulf.stroehler 2005-11-28 08:16:40 UTC
closing fixed/verified issue.