summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2012-03-13Version 3.4.6.2, tag libreoffice-3.4.6.2 (3.4.6-rc2)libreoffice-3.4.6.2Petr Mladek0-0/+0
2012-03-13fdo#46337: don't use an invalidated iteratorlibreoffice-3-4-6Luboš Luňák1-3/+4
(cherry picked from commit 2df1c40b4b7cb3107a68984db644f10097f04c3c) Signed-off-by: Michael Stahl <mstahl@redhat.com> Signed-off-by: Petr Mladek <pmladek@suse.cz> Signed-off-by: Norbert Thiebaud <nthiebaud@gmail.com>
2012-03-07Branch libreoffice-3-4-6Petr Mladek0-0/+0
This is 'libreoffice-3-4-6' - the stable branch for the 3.4.6 release. Only very safe changes, reviewed by three people are allowed. If you want to commit more complicated fix for the next 3.4.x release, please use the 'libreoffice-3-4' branch. Well, this is most likely the last 3.4.x bugfix release, so it probably does not make much sense. If you want to build something cool, unstable, and risky, use master.
2012-03-05fdo#42073: sw: expand all text fields when setting properties:Michael Stahl1-3/+2
SwXTextField::setPropertyValue: call Expand() for all text fields Fixes getPresentation returning stale values. (regression from CWS sw34bf01, 8485708f3001fca132c3353c464fe7187ef62bed) Also, without this the text formatting does not actually update and the new content is not displayed until some other unrelated editing operation. (cherry picked from commit e3dfae0741aae8581cd3fc713ba1b4459bb22d88) Signed-off-by: Miklos Vajna <vmiklos@suse.cz> Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
2012-03-05fdo#39694: SwTxtFld: do expand the field in the ctorMichael Stahl4-5/+11
While only the text formatting is able to expand all fields correctly, this is still good enough for many field types and fixes a race condition when the field value is requested before layout is finished. (regression from CWS sw34bf01, 8485708f3001fca132c3353c464fe7187ef62bed) (cherry picked from commit 9519deda120b73b72e75d89c3b2ae3d66220ec2d) Signed-off-by: Miklos Vajna <vmiklos@suse.cz> Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
2012-03-01SwRootFrm::~SwRootFrm: refactor:Michael Stahl5-5/+33
Since CWS swlayoutrefactoring the SwRootFrms are destroyed with SwDoc::IsInDtor not set. This can cause at least reads of freed SwRootFrm members when executing the dtors of SwRootFrm base class SwLayoutFrm calling into SwRootFrm::GetPageAtPos. Prevent this scenario by: - moving the implementation of SwRootFrm base class dtors to new methods SwFrm::Destroy and SwLayoutFrm::Destroy - calling SwFrm::Destroy and SwLayoutFrm::Destroy explicitly before SwRootFrm members are freed (cherry picked from commit ebb74441790a9852b1a1532d6e025c324666f6fc) Signed-off-by: Miklos Vajna <vmiklos@suse.cz>
2012-02-20fdo#39510: fix yet more layout crashes in ~SwRootFrm:Michael Stahl2-2/+2
Call SwRootFrm::RemoveFtns with the proper boolean to also eliminate EndNotes, and fix a faulty while loop in RemoveFtns that's been preventing removal of footnotes in certain sections since the dawn of CVS history. (regression from CWS swlayoutrefactoring, crashes on bugdoc from i#101776) (cherry picked from commit 962d0500c4debaef43e5f146e47e08c66d851562) Signed-off-by: Miklos Vajna <vmiklos@suse.cz>
2012-02-14fdo#45908: Cleaning up the tabs too early can cause loopsCédric Bosdonnat2-7/+2
Signed-off-by: Fridrich Štrba <fridrich.strba@bluewin.ch>
2012-02-13fdo#41712: sw: remove superfluous flag:Michael Stahl8-20/+3
This removes SwFlowFrm::bIsFollow, which is pointless duplication of m_pPrecede; this fixes lots of assertions (probably some real problem as well). Signed-off-by: Cédric Bosdonnat <cedric.bosdonnat.ooo@free.fr>
2012-02-13fdo#41712: sw: fix crash in layout frame linked lists:Michael Stahl4-2/+24
The pPrecede member is not maintained properly when setting the corresponding pFollow member. The change in SwTxtFrm::JoinFrm() fixes the crash, the other changes are perhaps fixes for other crashes... (regression from cc3d0d182cafef9649e45f4657233ac2221fdd0a) Signed-off-by: Cédric Bosdonnat <cedric.bosdonnat.ooo@free.fr>
2012-02-13fdo#41712: rename members to track down assignmentsMichael Stahl5-27/+29
Signed-off-by: Cédric Bosdonnat <cedric.bosdonnat.ooo@free.fr>
2012-02-07fdo#45115: SwXTextTable: fix setting bordersMichael Stahl1-1/+1
This plus e549f52f16c4a519ed3eddb9c66c19bacc247590 fixes the problem. Signed-off-by: Eike Rathke <erack@redhat.com>
2012-02-07fs34b: #i117545# (patch provided by mathias.bauer@oracle.com)Stephan Bergmann1-0/+5
Signed-off-by: Noel Power <noel.power@novell.com>
2012-02-07fdo#35661: fix sw.SwXViewSettings::com::sun::star::text::ViewSettingsBjoern Michaelsen2-3/+2
Signed-off-by: Noel Power <noel.power@novell.com>
2012-02-07fdo#37024: SwView::SwView: fix BROWSE_MODE setting:Michael Stahl1-1/+1
Apparently there is a check in SwDoc::SetAllUniqueFlyNames so that when frames with certain anchor types are in a document, then SetLoaded() is never called. Checking this flag via IsLoaded() here seems unnecessary. This problem was introduced with ebc5777548dea42ed966a16c66d879b1485bbfb4, from CWS swlayoutrefactoring. Signed-off-by: Noel Power <noel.power@novell.com>
2012-02-07fdo#38745: fix hilariously stupid stack guards:Michael Stahl1-2/+2
The UndoRedoRedlineGuards that SwUndo::{Un,Re}doWithContext wants to put on the stack aren't actually variables, so the destructor gets invoked before the function call that the guard is supposed to protect. Regression from CWS undoapi. (cherry picked from commit 13424b43c25389e303774c3fb2f2beb3e20ceae5) Signed-off-by: Stephan Bergmann <sbergman@redhat.com>
2012-01-30fdo#38542: sw: ODF import: prevent border width overriding:Michael Stahl1-16/+21
If there is a width in fo:border{,-left,-right,-top,-bottom}, then it should not override the values from style:border-line-width{,-left,-right,-top,-bottom}, which are more specific in case of "double" borders. Signed-off-by: Cédric Bosdonnat <cedric.bosdonnat.ooo@free.fr> Signed-off-by: Noel Power <noel.power@novell.com>
2012-01-30fdo#38542: sw: ODF import: divide width by 3 for "double" bordersMichael Stahl1-0/+4
The problem is that the width from the fo:border{,-left,-right,-top,-bottom} attributes is effectively tripled for "double" borders. Signed-off-by: Cédric Bosdonnat <cedric.bosdonnat.ooo@free.fr> Signed-off-by: Noel Power <noel.power@novell.com>
2012-01-27fdo#40438: force calculating layout before Activate to avoid crashes and loopsCédric Bosdonnat1-0/+4
(cherry picked from commit d83488f9795740857830aaf005e06e30d4e7d70c) Signed-off-by: Michael Stahl <mstahl@redhat.com>
2012-01-16fdo#38515: Fixed crasher in dialog destructorCédric Bosdonnat1-0/+3
Signed-off-by: Michael Meeks <michael.meeks@suse.com>
2012-01-02fdo#36874: Label PRINTING misplaced on paper sheetWinfried Donkers1-3/+7
Signed-off-by: Petr Mladek <pmladek@suse.cz>
2011-12-22sw: fdo#39159 fdo#40482: temp selection print doc:Michael Stahl7-42/+47
Ensure that the printing temp selection document is not destroyed prematurely by SwXTextView::NotifySelChanged, called via ViewOptionAdjustStop, by retaining the temp doc object shell not at the View but in SwRenderData. Not restoring the view options for selections does not actually work, because having a selection surprisingly does not imply printing a temp document: the preview also uses a selection. (view option regression from cd690d2e72be410058376c416a40ff5d918fb0f7) backport of 89d2733e16ae6233deea6bef3193bd45c89b854c Signed-off-by: Caolán McNamara <caolanm@redhat.com>
2011-12-13docx: Back-port Lubos' comment export workMichael Meeks7-7/+138
Signed-off-by: Michael Stahl <mstahl@redhat.com>
2011-12-09fdo#40831: SwView::ExecSearch:Michael Stahl1-2/+15
Replace may delete nodes, so the current cursor position must be stored in a way so that it is corrected when nodes are removed. Signed-off-by: Michael Meeks <michael.meeks@suse.com>
2011-12-09i#102333: SwDoc::ReplaceRangeImpl:Michael Stahl1-2/+5
The fix for this (230fcf4a456636bb466f72834cd57238621d206d) was not quite right; there are 2 different join modes, and moving the cursor backward is required in one of them, but not the other. This can be seen when doing replace all ^$ with several consecutive empty lines. Signed-off-by: Michael Meeks <michael.meeks@suse.com>
2011-12-07backport of 3524727db0f3cfecf3a47046795c527808c10c3eLuboš Luňák4-10/+17
commit 3524727db0f3cfecf3a47046795c527808c10c3e Author: Caolán McNamara <caolanm@redhat.com> Date: Thu Jun 23 15:14:00 2011 +0100 Related: #i58612# don't crash anyway
2011-11-10NO_STYLE default for borderlines mso import, fixes image size issue bnc#718971Noel Power1-1/+1
Signed-off-by: Michael Meeks <michael.meeks@suse.com>
2011-10-26Fix fdo#41995 fallout - recognize .svg in odt containerThorsten Behrens1-5/+5
Seems the graphic load code is stupid and not using the path name / file extension to guess file type, but only "magic byte" detection. Giving filter framework the path now, so that .svg actually loads. This parrots 050350f326629c6edffcdb2cf288893027f264e6, but for Writer, which sadly sports its own Graphic implementation. Signed-off-by: Noel Power <nopower@suse.com>
2011-10-25Version 3.4.4.1, tag libreoffice-3.4.4.1 (3.4.4-rc1)libreoffice-3.4.4.1Petr Mladek0-0/+0
2011-09-26fdo#39510: comment addedBjoern Michaelsen1-0/+10
Signed-off-by: Michael Stahl <mst@openoffice.org>
2011-09-26fdo#39510 lp#854626: crash on closing document with footnotesBjoern Michaelsen1-0/+1
Signed-off-by: Michael Stahl <mst@openoffice.org>
2011-09-21Fix for fdo#35513: avoid crash while processing incorrect range of pagesIvan Timofeev1-1/+4
Signed-off-by: Michael Meeks <michael.meeks@novell.com>
2011-09-07Fix leak in pdf export.Andor Ertsey1-0/+2
This fixes i#116448 - SwViewOption was allocated but never freed during pdf export - killed mass document exporting via OOM eventually. Signed-off-by: Caolán McNamara <caolanm@redhat.com>
2011-08-24fdo#40341: hide 'math baseline alignment' unless availableTakeshi Abe1-5/+11
Signed-off-by: Caolán McNamara <caolanm@redhat.com>
2011-08-23cherry-picked core:cc3d0d and fixed a few thingsCédric Bosdonnat6-10/+60
2011-08-16Version 3.4.3.1, tag libreoffice-3.4.3.1 (3.4.3-rc1)libreoffice-3.4.3.1Petr Mladek0-0/+0
2011-08-10Resolves: fdo#39159 don't restore original view settings on temp documentsCaolán McNamara4-8/+15
2011-08-10put exception message into mailmerge error dialogCaolán McNamara1-3/+8
Signed-off-by: Noel Power <noel.power@novell.com>
2011-08-09Fixed invalid pointer dereferenceMarc-Andre Laverdiere1-3/+3
(cherry picked from commit 6f32f7f8896dcb21e983adcb61ed400db7389a34)
2011-08-09Fixed invalid array bounds read in ww8scan.cxxMarc-Andre Laverdiere1-1/+2
(cherry picked from commit a1c11fa56aef80b4a9a47054029613fb06a30832)
2011-08-09Fixed valgrind errorMarc-Andre Laverdiere1-6/+7
Complained about switching on an unitialized value (cherry picked from commit 5d7caec204e704c62d670cc90a39e8cfc78479c6)
2011-08-09Fixes for segfault on an edge case + translationsMarc-Andre Laverdiere1-16/+14
Fixed segfault due to reading an invalid pointer returned by GetStyle Translated some of the comments Added some proofing against similar bugs from GetStyle that may happen in other edge cases Fixed compile warning (cherry picked from commit b2f9a798a010f76e887a7f10c6f0af667fe48e81)
2011-08-09Fixed invalid array dereferenceMarc-Andre Laverdiere1-14/+15
(cherry picked from commit 560b004d9812a76783c701724043f6cee70c0928)
2011-08-04Resolves: fdo#32899 backport fixos5-2/+41
2011-08-03Related: #i38880# crash in new field codeCaolán McNamara1-2/+5
Signed-off-by: Noel Power <noel.power@novell.com>
2011-08-01Resolves: fdo#37693 it really is a LineBorder not a LineBorder2Caolán McNamara1-13/+8
Signed-off-by: Luboš Luňák <l.lunak@suse.cz>
2011-07-27Resolves: fdo#36631 don't crash on undo of ole2 insertCaolán McNamara1-1/+6
Signed-off-by: Jan Holesovsky <kendy@suse.cz>
2011-07-26Resolves: #i118018# fExtChar affects only < 8 (if at all)Caolán McNamara1-19/+21
2011-07-18starmath: check for edit window when inline editing is enabledMiklos Vajna1-1/+2
Signed-off-by: Michael Meeks <michael.meeks@novell.com>
2011-07-18starmath: check for existing edit windowMiklos Vajna1-0/+2
Signed-off-by: Michael Meeks <michael.meeks@novell.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-18 02:24:40 +0000