-
Halla Rempt authored
Merged revisions 1112881,1112927,1112977-1112982,1112989,1113002,1113005-1113008,1113045,1113049-1113050,1113218 via svnmerge from https://rempt@svn.kde.org/home/kde/trunk/koffice ........ r1112881 | berger | 2010-04-09 14:17:17 +0200 (Fri, 09 Apr 2010) | 6 lines Make KoColor::setPixel takes a const pointer (and also reallocate memory only if necesserary) Make KisPaintDevice::pixel be const and use const iterator, this avoid the memento manager to start creating a new history point, and therefor fix the KisTransaction test ........ r1112927 | lassin | 2010-04-09 15:59:00 +0200 (Fri, 09 Apr 2010) | 3 lines Fixes cases where paragraphs has both par and text properties in styles.xml ........ r1112977 | zander | 2010-04-09 17:39:27 +0200 (Fri, 09 Apr 2010) | 1 line Remove redundant cast ........ r1112978 | zander | 2010-04-09 17:39:29 +0200 (Fri, 09 Apr 2010) | 1 line Fix tab offset with parag that has indent. ........ r1112979 | zander | 2010-04-09 17:39:31 +0200 (Fri, 09 Apr 2010) | 1 line Fix background color ........ r1112980 | zander | 2010-04-09 17:39:33 +0200 (Fri, 09 Apr 2010) | 1 line Change default when XML does not specify it to noWrap ........ r1112981 | zander | 2010-04-09 17:39:35 +0200 (Fri, 09 Apr 2010) | 3 lines Fix NaN creation on height or width zero. BUG:231846 ........ r1112982 | zander | 2010-04-09 17:39:37 +0200 (Fri, 09 Apr 2010) | 1 line Add some asserts ........ r1112989 | lukast | 2010-04-09 18:09:16 +0200 (Fri, 09 Apr 2010) | 4 lines Revert "Fix the expected value in the test." Seems like magic stuff happens. Now it is 7 and the test passes. I'm not aware what might change. ........ r1113002 | sebsauer | 2010-04-09 18:48:38 +0200 (Fri, 09 Apr 2010) | 2 lines Fix for the parser-issue discovered by Carlos. We cannot just keep the whitespaces even if removing them is wrong too cause compile() does lot of compression and token modifications which do not proper handle whitespaces. Seems what needs to be done is to change the meaning depending on the referenced content and we cannot do that on compile() but only on evalRecursive(). That in turn means that compile() needs to be extended to transparently ignore operators so we have them still available on evaluation to decide then what to do with them. For now let's go with the easiest solution and just ignore (as in remove) the intersect operator. Lot of thanks to Carlos :) ........ r1113005 | berger | 2010-04-09 19:34:44 +0200 (Fri, 09 Apr 2010) | 4 lines better delete the node model before the image (since the node model is referencing the layers themselves) ........ r1113006 | berger | 2010-04-09 19:36:32 +0200 (Fri, 09 Apr 2010) | 3 lines disable the projection, it is not what we are testing in the tests and it is causing random crashes in the tests ........ r1113007 | berger | 2010-04-09 19:46:57 +0200 (Fri, 09 Apr 2010) | 2 lines since painterly framework is not installed, disable its test ........ r1113008 | sebsauer | 2010-04-09 19:50:47 +0200 (Fri, 09 Apr 2010) | 3 lines Make POINT_TO_MM & co return qreal's http://reviewboard.kde.org/r/3526/ ........ r1113045 | berger | 2010-04-09 21:13:51 +0200 (Fri, 09 Apr 2010) | 7 lines Fix: change the color of the solid color generator result in a black generator Stop using legacyXML BUG:233796 ........ r1113049 | berger | 2010-04-09 21:27:33 +0200 (Fri, 09 Apr 2010) | 2 lines Remove last bit of toLegacyXML ........ r1113050 | berger | 2010-04-09 21:29:57 +0200 (Fri, 09 Apr 2010) | 2 lines --warnings (int vs uint) ........ r1113218 | cfeck | 2010-04-10 05:00:37 +0200 (Sat, 10 Apr 2010) | 6 lines SVN_SILENT compile after r1113049 Please verify :) CCMAIL: cyb@lepi.org ........ svn path=/branches/work/koffice-ko/; revision=1113251
94a77582