summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2021-09-23tdf#144624 Prevent infinite recursion when loading SVG iconAleksei Nikiforov
When Generic/X11 VCL backend plugin loads SVG icon, it creates virtual device for rasterizing an SVG icon, which in turn tries to load an SVG icon, and thus infinite recursion happens. Change-Id: I7559b6255e6718e64ef4a6e7c79d597375e5823a Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122344 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 9b4478e62c712ef0f75c4a001e260dfdd6b3ca4c) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122518
2021-09-23tdf#143722 DOCX import: map using default TOC heading styleGabor Kelemen
of Writer ("Contents Heading") instead of "TOC Heading". This essentially reverts the fix to bug #83300. When that was added the TOC import was broken enough that it made no visible regression. In 6.3 the TOC import was fixed up so now it's more annoying. The unit test of bug #83300 is reused but depends on the fix to bug #143726. I checked the example file of bug #83300 and the layout problem of comment 6 that was fixed there does not come back. Follow-up to commit 5440492ff9f949ee9ed9052e8bab6f5136d78b2a "tf143726 DOCX: export default TOC Header style with correct name". Change-Id: Iaee73729ea46a0c36d08ccc6fc5fb04fdf592a1d Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120092 Tested-by: László Németh <nemeth@numbertext.org> Reviewed-by: László Németh <nemeth@numbertext.org> (cherry picked from commit e3f9170c03eb9121b0c244c0e2e60d15e6920deb) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122279 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-23tdf#144631 Colibre: sync up query/sql/report execution iconsRizal Muttaqin
Change-Id: I52cfa4aa52f0bc03accb1030d463893ee871cdab Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122504 Tested-by: Jenkins Reviewed-by: Rizal Muttaqin <rizmut@libreoffice.org> (cherry picked from commit 0b810cc0f04241e644c82ba8bdad6a075b964118) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122435 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-23Resolves: tdf#142704 %PRODUCTNAME shown in gtk3 extended tipsCaolán McNamara
Change-Id: I61d8b83ce326816c498f54e3cfc053270d82c1a3 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122433 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-22tdf#144091 svx: fix unwanted blur of shadow from table cell fillMiklos Vajna
Initial render support for shadows of table shapes were added in commit a75bf43a8d6c5dec6dcc86908c142ceec541aa8c (tdf#129961 svx: add rendering for table shadow as direct format, 2020-12-02). That already noticed a trick with the shadow of table shapes: the shadow is generate from the cell fill and the border, but not from the text. An additional trick is that when blur is enabled for the table shape's shadow, then only the border should be blurred, not the cell fill. In the bug document's case, the effective cell background was gray, with a semi-transparent red shadow. We used to render cc0000 with blur and cccccc without blur, now we correctly render cca3a3, matching PowerPoint. (cherry picked from commit 37a52d30bbfcf1d073779b50139c4dafa507be4b) Conflicts: drawinglayer/source/primitive2d/shadowprimitive2d.cxx drawinglayer/source/tools/primitive2dxmldump.cxx include/drawinglayer/primitive2d/BufferedDecompositionPrimitive2D.hxx Change-Id: I7326a5f6254cf19b2d05181084c78e734ff7a7b4 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122357 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
2021-09-22tdf#127900 sd: fix regression at applying master propertiesAttila Bakos (NISZ)
Language settings of frames in slide master weren't applied on the slides based on that master. Regression since LO 6.3 by "tdf#126067 Fix slide scope feature." (commit 40bb9ac690d979ef544d5aa759bd734a176912a0). Co-developed-by: Dániel Arató (NISZ) Change-Id: I559adbe00870ed8a3d2947fef8dae435a387e34a Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120993 Tested-by: László Németh <nemeth@numbertext.org> Reviewed-by: László Németh <nemeth@numbertext.org> Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122397 Tested-by: Jenkins
2021-09-21bump product version to 7.2.3.0.0+Christian Lohmaier
Change-Id: Iccb47e7daed8206882f14d48454fc46129466801
2021-09-21Update git submodulesChristian Lohmaier
* Update translations from branch 'libreoffice-7-2' to e6c8a0bddf2aa2983c8a75b48b9802319bd0c88f - update translations for 7.2.2 rc1 and force-fix errors using pocheck Change-Id: I64567c3dbe7e63de2550530fb5e0e0ba8e7f0525
2021-09-21tdf#144253 tdf#144324 sc filter: use formatted values in filtersBalazs Varga
if the results of formulas are values. Followed up of 40acda4e78127fa9f513646ef210b074d40cf307 (Related: tdf#140968 avoid duplicated filter values) Change-Id: Ib396d2b7cc08ba41b5936a53a28b5e38bf678b3e Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121715 Tested-by: László Németh <nemeth@numbertext.org> Reviewed-by: László Németh <nemeth@numbertext.org> (cherry picked from commit 51375b48378915b6e95c1ac26b2ccf8e39880f7e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122288 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-21tdf#143591 DOCX import: handle anchored objects as at-charMiklos Vajna
This partially reverts cc8f8ae55f681755f5da3bf64e4c30bb713f0383 (DOCX drawingML shape import: wp:anchor's behindDoc attribute, 2013-11-19), it seems to be more important to be consistent with the DOC import than with the VML import which is no longer used for DOCX shapes crated with Word >= 2010. Change-Id: I631da010bce1b4d3c392645e0ae3797a03665a42 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122367 Tested-by: Jenkins Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 7246e57216bb20c15af0ecf6a0183f5ffa81e780) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122287 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-21Resolves: tdf#144564 don't query siblings and parent to determine typeCaolán McNamara
because this is used during sorting of the tree elements so its position isn't necessarily meaningful during the sort. DBTreeListUserData is supposed to exist for elements not staged for removal and that already has the type as a member Change-Id: Ie1004dbcdca2fae8711941d98a084103a0b15815 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122355 Reviewed-by: Michael Stahl <michael.stahl@allotropia.de> Tested-by: Jenkins
2021-09-21tdf143726 DOCX: export default TOC Header style with correct nameGabor Kelemen
This seems to have been a typo since it was mapped to the default name of the Table of Authorities index's heading in Word which is not really supported anyways. Change-Id: I4cadce18c30c5497f27479fcc251fdf85d859145 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/120091 Tested-by: László Németh <nemeth@numbertext.org> Reviewed-by: László Németh <nemeth@numbertext.org> (cherry picked from commit 5440492ff9f949ee9ed9052e8bab6f5136d78b2a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122278 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-21use clang-cl's -Zc:dllexportInlines- for Skia (tdf#144598)Luboš Luňák
This is clang-cl's equivalent of -fvisibility-inlines-hidden, and it seems to be also sort of the equivalent of MSVC's -Zc:inline. So it saves build time and disk space. As an additional effect, this disables emitting copies of inlines functions in every .o file where the function is called (even if inlined), which means that it hopefully avoids the problem of SkOpts_avx.cpp generating a copy of SkRect::round() which would include AVX code, and the linker might select this as the instance of SkRect::round() to keep, thus making SSE2 code call AVX code without checking for AVX availability first. Change-Id: I97541ae11d05f489894bc9233271eb21fd520f43 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122335 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit 36f76223193fb96df7b8cbc1a1ff30f739857189) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122284 Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-21Revert "tdf#99492 sw layout: always paint merged cell grids"Justin Luth
This reverts LO 7.1 commit c016fe2b5918d6e53113e100b1126076b6e1a6a3. This failed for two reasons. 1.) If it is not a solid color (like a graphic for example) then the graphic is going to be repeated in each cell. 2.) This is NOT actually repainting the visible cell's background. It is painting the hidden cell - which MIGHT be different from the visible cell (in the case of a solid color). Since this will require a completely different approach, I am just reverting this. Thanks Mike for finding an example document that shows the flaws. Change-Id: Icdc21e09118e7c33ac9f7ede23c913b95ad69c93 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122366 Tested-by: Jenkins Reviewed-by: Justin Luth <justin_luth@sil.org> (cherry picked from commit eb830ad284f245165b6ab5e8647d48834622f2d5) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122281
2021-09-21Resolves: tdf#144620 toolbar config item overwrites options dialog oneCaolán McNamara
There are two SwMailMergeConfigItem involved. One SwMailMergeConfigItem from the toolbar created and considered modified and with changes not commited to the config yet. Then the other SwMailMergeConfigItem is created by the options dialog and commited on ok, but after that commit at cui/source/options/treeopt.cxx at line 723 there is a utl::ConfigManager::storeConfigItems() to flush all outstanding config items, so the one belonging to the toolbar is now flushed after the options dialog one was written. The SwMailMergeConfigItem has a IsModified() of true right after ctor which doesn't seem intentional, there is no explicit set of Modified to true on setting the simple bool members, the Modified bit is toggled on when using the more complicated modifier methods during ctor so very much looks like an accidental side effect. Change-Id: If84a6f01c7bf92704dd1e175a2bd8e2ea59e157f Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122280 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-20tdf#143813 Outline folding: Make copy cut paste workJim Raykowski
Change-Id: Ib8b491fdcce1cd059c8eaf80a9c3bb2590af7c63 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122212 Tested-by: Jenkins Reviewed-by: Jim Raykowski <raykowj@gmail.com> (cherry picked from commit f4f646eec90c44b6a7ffeaf5f6ce4c85af64eed0) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122276 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-17occasionally the bg color of the deck grip is wrongCaolán McNamara
Change-Id: I318e817a8e99b9f01aee90f042bfac39a1c4b5d9 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122120 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-17tdf#144427 tdf#144483 sw mailmerge: fix merge all documentBalazs Varga
Not all records were merged if we selected the “From 'X' To 'Y' of the “Save Merged Documents” or “Print Merged Documents” or "Send to Email" dialog windows and write some value into them and then we switched back on to Save/Print/Send All record (opened via the Mail merge toolbar). Regression from commit: ec44f87d5b99a3299322d0b79abc4c6808877865 (tdf#117212 sw mailmerge: merge only the selected range) (cherry picked from commit f3993912ec4b526aa57cb4bfb4745d7298a4da82) tdf#144483 sw mail merge: save ranges in a single document In the Save mail merged document dialog, it wasn't possible to save mail ranges in a single document, because of the bad UI of the dialog window: grouping range selector as a radio button with the single document/individual documents radio buttons. Moreover, commit f3993912ec4b526aa57cb4bfb4745d7298a4da82 "tdf#144427 sw mailmerge: fix merge all document" removed the hidden workaround: setting range at the third radio button, and selecting the first radio button. Using checkbox for the third option solved the problem as proposed. (cherry picked from commit 45c4caff685b15a0f1b87ef05436a7e6aca96851) Fix broken ui file ...after 45c4caff685b15a0f1b87ef05436a7e6aca96851 "tdf#144483 sw mail merge: save ranges in a single document", causing CppunitTest_sw_dialogs_test to fail with > warn:vcl.gtk:2759211:2759211:vcl/unx/gtk3/gtkinst.cxx:21611: GtkInstanceBuilder: error when calling gtk_builder_add_from_file: ~/lo/core/instdir/share/config/soffice.cfg/modules/swriter/ui/mmresultsavedialog.ui:165:49 Invalid property: GtkCheckButton.group (cherry picked from commit 7b37af5af6afe75ad952538c145a4f4e61de9a96) Change-Id: I01fc664fe76f74cefe4faa81b324088ec37b9881 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121982 Reviewed-by: László Németh <nemeth@numbertext.org> Reviewed-by: Balazs Varga <varga.balazs3@nisz.hu> Tested-by: Jenkins
2021-09-17Resolves tdf#144523 - Update Liberation fonts to 2.1.4 releaseHeiko Tietze
Change-Id: I25aa71d214eec3a131a7b11dfe292764061a127c Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122182 Tested-by: Jenkins Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org> (cherry picked from commit 95406f2e172df13ae9c6536ca1378cd7fca3b2de) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122129 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-16Revert "speedup rational_FromDouble"Noel Grandin
This reverts commit 67d83e40e2c4f3862c50e6abeabfc24a75119fc8. because it causes less accuracy sometimes, eg. see https://gerrit.libreoffice.org/c/core/+/122186 Change-Id: I06eb4921359a8fe6689c6cfd3e2967e25b646fa3 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122124 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk> (cherry picked from commit f2315daf701847a180a5759039b4951fb4da0653) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122126
2021-09-16CppunitTest_sw_core_text needs some more RTTI now in UBSan buildsStephan Bergmann
...presumably since 0c3e47cc2f0570a7cd8ff4889763991a86b29f26 "tdf#144305 sw: fix rendering of ruby portions with non-default ruby alignment", > DynamicLibraryManagerException: "Failed to load dynamic library: ~/lo/core/workdir/LinkTarget/CppunitTest/libtest_sw_core_text.so > ~/lo/core/workdir/LinkTarget/CppunitTest/libtest_sw_core_text.so: undefined symbol: _ZTI13SwLinePortion" and > DynamicLibraryManagerException: "Failed to load dynamic library: ~/lo/core/workdir/LinkTarget/CppunitTest/libtest_sw_core_text.so > ~/lo/core/workdir/LinkTarget/CppunitTest/libtest_sw_core_text.so: undefined symbol: _ZTI14SwMultiPortion" Change-Id: Iaebffc3c9c0245d6092ef1dfc58213fd2d38d303 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122143 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit 485add3cd6240a0faceb7a74784d44b1c0a6705e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122117 Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-16patch Skia to search only system32/ for librariesLuboš Luňák
LO on Windows defaults to Vulkan, and Skia tries to load the vulkan DLL, which may not be present on some systems. The library AFAICT should always be in a system directory, so restrict library searching to just there to avoid the possibility of DLL hijacking, just to be safe. Change-Id: I78ad3c7297e613c0316e82c5ff3c0110a02da337 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121437 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122135 Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-09-16fix skia build with newer freetypesRene Engelhard
extracted (and adapted for 7.2.x) from masters catch-all-commit a0edcc68f94915a78fcc08e70d2cdd752abd9ebb: Additionally patch out Skia's use of TT_SUPPORT_COLRV1, which seems to be an unstable freetype API from the git version and it doesn't even compile with the latest stable 2.9.11 release Change-Id: Iba22fbc74dcd75bc6d1d91e2f537caf9d179e885 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122096 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
2021-09-15Resolves: tdf#144135 Rework Sort with area extrasEike Rathke
Since commit 774a61afa9fc281290e7bfad4e28c05978b82d73 CommitDate: Wed Apr 14 08:46:03 2021 +0200 tdf#126678 - Consider "Include formats" option during sort a sheet formatted with visible attributes like cell background colour up to the end if for Sort all columns and/or rows are selected lead to an excessive memory allocation and slow execution time if it didn't get killed by the operating system before due to memory exhaustion. The same could had happened already before if graphics or comments were to be included that could had resulted in a similar large range. However, cell formats across sheets are more likely. This changes the strategy how the to be sorted data range is determined (range only with data) and additional area extras ranges without data that are only to be rearranged. Those are then processed in chunks (limited to ~512MB per chunk). Cell formats that are identical within one column's rows range do not even need to be covered as they are not rearranged, in the best case leading to all trailing formats' ranges being excluded from the sort. Additionally optimize the cell gathering of formats, graphics and comments such that for the area extras they are only collected if actually requested. The overall performance gain is in an order of magnitudes even if some extras are to be collected. Change-Id: If3abbaeaa615aaff7d88a82a5b3fc7ac633d770d Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122013 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 0a9b68c9f9880655576e3220d8b70064b367dbee) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121981 Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
2021-09-15Update git submodulesOlivier Hallot
* Update helpcontent2 from branch 'libreoffice-7-2' to f3bf3ced8ac07689e83c4f29226323313a23caed - tdf#133851 connect help page to certificate path dialog Change-Id: I5a96785f32b81c6a8dd347b5665e1f5d0c7b84b2 Reviewed-on: https://gerrit.libreoffice.org/c/help/+/122152 Tested-by: Jenkins Reviewed-by: Olivier Hallot <olivier.hallot@libreoffice.org> (cherry picked from commit 323c851c00559631e9e2ee1098d9aed89ca1d2ff) Reviewed-on: https://gerrit.libreoffice.org/c/help/+/122118 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-15Adapt to Bison 3.8 internal yyn -> yyrule renameStephan Bergmann
see <https://git.savannah.gnu.org/cgit/bison.git/commit/?id=f30067ed51f23802fc91761ede1506dfa72b2865> "glr2.cc: log the execution of deferred actions" including "Rename argument yyn as yyrule for clarity." YYBISON was defined as 1 rather than as a representation of the Bison version prior to <https://git.savannah.gnu.org/cgit/bison.git/commit/?id=21c147b6e5372563b7c4741deadaddb9354f4b09> "yacc.c: provide the Bison version as an integral macro", which shouldn't be a problem here. And YYBISON is apparently completely undefined with /usr/bin/bison on macOS. (The preceding comment always mentioned "yyi" and "yyrmap" in apparent mismatch with the actually used "yyn" and "yyr1" ever since c25ec0608a167bcf1d891043f02273761c351701 "initial import", so just leave it untouched.) Change-Id: I4f901407aa21ed4abec84e661d813ee7599f02f0 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122082 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit 45227d9b79dc4f2a2aa6874cd4e3c02b7934b197) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122069 Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-14tdf#144305 sw: fix rendering of ruby portions with non-default ruby alignmentMiklos Vajna
Regression from 301278b656e76b6f42af5cf8a6f5c6c02acfffeb (sw: allow the height of a line to be larger than 65536 twips, 2021-05-20), the problem was that changing from sal_uInt16 to sal_uInt32 broke SwRubyPortion::Adjust_(), which relied on integer promotion rules to have a negative diff. Old storage size was smaller than int, so got promoted to signed int, so the result could be a small negative number. New storage size is an unsigned int, so no promotion happens, so the new result was a large positive number. Fix this by casting to signed int explicitly. Change-Id: I8778c1bd0d62e27c99d4ceb1bb7bc6107a179803 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122048 Tested-by: Jenkins Reviewed-by: Miklos Vajna <vmiklos@collabora.com> (cherry picked from commit 0c3e47cc2f0570a7cd8ff4889763991a86b29f26) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122066 Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-14tdf#142415 mouse events not propogated to table control event handlersCaolán McNamara
handle this with explicit callbacks from the cell widget for those events Change-Id: Ie605ca4286afc0fbd321f339fb7963771a303df5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122063 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-14tdf#117842: use XUnoTunnel to get wrapped connection reliablyMike Kaganski
XDataDefinitionSupplier::getDataDefinitionByConnection may take not only direct reference to expected connection type, but also connectivity::OConnectionWeakWrapper wrapping it. Change-Id: I88925f9403b51f0cf13f02b5f00d3765a242230e Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121890 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> (cherry picked from commit ccc6c846eb7f6d334d0ab2e6b50c188c434caa19) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121986 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-14tdf#123642: keep last bookmark at the document endVasily Melenchuk
In some cases DomainMapper_Impl::RemoveLastParagraph() can also remove last bookmark from real last paragraph. This does never happens when we use xParagraph->dispose(), but pretty always during older way with xCursor->setString(OUString()). Unfortunately without deep refactoring of redlines, bookmarks, etc. I see no other way to avoid this removal except given hack which is trying to store last bookmark and if it did disappear restore it. Some existing unittests were adjusted: corresponding original DOCX files did contain final bookmarks not taken into account by the code. In some cases test code is modified, in some just removed final bookmark in DOCX: such multi-format tests as in ooxmlfieldexport.cxx will be more identical to RTF & DOC variants. Conflicts: sw/qa/extras/ooxmlexport/ooxmlexport17.cxx Change-Id: Ie9948b58cda705a0b85fa8e5e08b72fbb7d682b2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121409 Tested-by: Jenkins Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122015
2021-09-14tdf#144106 Don't proceed ptrim_i past ptrim_fStephan Bergmann
...which would cause p2 - p1 to be negative in lcl_appendLineData and thus construction of a std::u16string_view with a huge positive length of type size_t. In 64-bit builds where size_t is 64-bit, that would then cause termination due to an uncaught std::bad_alloc. But in (implicitly) --disable-assert-always-abort 32-bit builds where size_t is 32-bit, this would silently have worked before 1efec9ec21dba32335e311d367b636538e219621 "Tighten rtl_{string,uString}_newFromStr_WithLength implementation", when the huge positive size_t value was cast back to a negative sal_Int32 that was gracefully handled by rtl_uString_newFromStr_WithLength. Change-Id: I3b95a9fce62b99ffc150f76a1c6ccddcdacdae0e Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122038 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit 4a4be7a1edead11b48e1a8598e52a3246e6744bb) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122064 Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-09-13do not intern a temporary cell error stringLuboš Luňák
Interning is way more work than a simple text comparison done once. Change-Id: If18c478fc62d1fb09ce2141fdb77b46a6bc46c08 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121855 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit d0316985db22efd6708dffa173eaabb430f6b9a8) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121985 Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-09-13exernal/firebird: extern/cloop: Missing dependencies of compilations on...Stephan Bergmann
...output directories Change-Id: Iaa2b750a12e3df078b46e5bb4feeafc926e11165 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121741 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122041 Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-09-13Explicitly set cursor position for x11 'gen' backend, tdf#144457 follow-upEike Rathke
... (and other backends?) Change-Id: If45b83080aa2df50ef27ad282eb6fa1d4a022703 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122037 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 37e2e99f7a3018dce0337b582b139d41e1e81a9a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122059 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-13Resolves: tdf#116184 Check that there is no trailing number behind a matchEike Rathke
... without being separated by a blank so the match is rejected if it doesn't possibly form a date+time input and input can be accepted as decimal fraction. Change-Id: Iabd1d216366ecb8454c59822ce58f112bfa6091e Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122024 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit eb0b4ab2d3b86d77ee0edb652d4486343e5b3b1f) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122054 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-13cache FormulaError::NoRef error string (tdf#144249)Luboš Luňák
Those COUNTIF() may call it a huge number of times, and the translation of the string each time then actually is a noticeable impact. And ScGlobal already does one-time intialization of objects based on the locale, so one-time initializing a string there should be fine too. Change-Id: I0daeb50ccb43f780d99cf3838cfd1bf790c5f6cc Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121856 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit b4c1bb2f91e9ae47820c289e2c08f640a958cf05) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121984 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-13tdf#115547: Fix DB path handlingMike Kaganski
1. The code used wrong procedure to convert file URLs to local paths. It assumed that it's enough to just strip the leading 'file://', which is only sometimes true on Unix-like systems; on Windows, this converts a valid 'file:///C:/path/file.ext' to '/C:/path/file.ext', where the leading slash is then treated as a network path, resulting in errors. 2. It is incorrect to assume the same length for UTF-16 and UTF-8 encoded URLs coming from untrusted source (like user file). It may contain non-ASCII characters (be an IRL), and then the assumption would fail. Change-Id: Ie2ea6c8cb9a690975db956fa025bf926a8010984 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121885 Reviewed-by: Lionel Mamane <lionel@mamane.lu> Tested-by: Jenkins (cherry picked from commit 51269c4d28c04ebd2c0047772b7373e0bebec219) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121983 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-13Related: tdf#144389: merge ThumbnailView and ThumbnailViewBaseCaolán McNamara
only ThumbnailView inherits directly from ThumbnailViewBase and it will simplify a11y if ThumbnailView instead of ThumbnailViewBase is available Change-Id: I715faa3f9b2cec68c1de07479b7d1dbbd9ddbcc2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121848 Tested-by: Jenkins Tested-by: Caolán McNamara <caolanm@redhat.com> Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-09-13tdf#144352 Revert "Fix widget placement problem on mailmerge result dialog."Gabor Kelemen
This reverts commit 0a3e5ba4db254549c15e55984f1764a352c8239b. Reason for revert: Causes the dialogs bottom part to slip on top each other, most visibly on Windows Change-Id: I6e5a30fbb4611a7b1250b173adccdb47c95eea02 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121879 Tested-by: Jenkins Reviewed-by: László Németh <nemeth@numbertext.org> (cherry picked from commit c10c32438025edda2f3c178b47d369e7979f25a8) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121988 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-13Resolves: tdf#144389 get_accessible_parent should only be called on vcl implsCaolán McNamara
Change-Id: If5d8718fcd9bcccee37e162a99cc68ff4a77de8f Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121849 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-13crashtesting: threaded scaling crash on re-export of ooo24840-1.sxw to odtCaolán McNamara
#13 0x00007f1cb843752a in o3tl::cow_wrapper<ImplBitmapPalette, o3tl::UnsafeRefCountingPolicy>::operator->() (this=0x5596086d5968) at include/o3tl/cow_wrapper.hxx:329 __PRETTY_FUNCTION__ = "BitmapColor& BitmapPalette::operator[](sal_uInt16)" #14 0x00007f1cb843752a in BitmapPalette::operator[](unsigned short) (this=0x5596086d5968, nIndex=nIndex@entry=0) at vcl/source/bitmap/bitmappalette.cxx:139 __PRETTY_FUNCTION__ = "BitmapColor& BitmapPalette::operator[](sal_uInt16)" #15 0x00007f1cb849f5f5 in BitmapInfoAccess::GetPaletteColor(unsigned short) const (nColor=0, this=0x5596085989f0) at include/vcl/BitmapInfoAccess.hxx:114 __PRETTY_FUNCTION__ = "const BitmapColor& BitmapInfoAccess::GetPaletteColor(sal_uInt16) const" the mpBuffer member of BitmapInfoAccess is BitmapBuffer* mpBuffer; not const BitmapBuffer* mpBuffer; so mpBuffer->maPalette.foo() calls non-const variants of foo(), (BitmapPalette::operator[](unsigned short) in this case), which is presumably non the expected outcome, as the copy-on-write mpImpl of BitmapPalette unsafely creates a new copy its internals on the first dereference of mpImpl in a non-const method. Change-Id: I1ebb3c67386a9028e5b8bab4b2d1cc5862700aa1 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121888 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2021-09-13tdf#144425 Set sane lower bounds (1) for MM e-mail output rangesGabor Kelemen
Lower bounds were forgotten to be set in weld commit f1ca64800074530d95e507f93c764a687310b9eb for the new GtkAdjustments of the GtkSpinButtons This caused no visible differences until commit d9fa826769cd570814f3556d53493a78d2869873 when new default values (0) were added for VCL FormattedFields This made it possible to email MM results on non-GTK vclpugs starting from 0th mail if custom range is chosen, which causes an instant hang in the sending process since there is no -1st generated result. The default Send all option has still worked after this. Then commit ec44f87d5b99a3299322d0b79abc4c6808877865 started to use the default GtkSpinButton values for default range of result generation, breaking the Send all option as well. Change-Id: I2a9f2b0954045700f947f342e5928ef75ce23aed Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121865 Reviewed-by: László Németh <nemeth@numbertext.org> Tested-by: László Németh <nemeth@numbertext.org> (cherry picked from commit e85aaaecb5479660aa0cf600564ee3caa470aa3d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121877 Tested-by: Jenkins
2021-09-13tdf#144412 Qt5 reset m_nModKeyCode on key inputJan-Marek Glogowski
Switching the input direction on Ctrl + L/R-Shift is actually no bug but a feature. It is triggered on key release, so it can be distinguished from / doesn't interfere with shortcut handling. That's what should happen. So trying to implement that behaviour correctly and seeing the appearingly wrong modifier events for gtk3, I found gtk3 resets the frames persistent mnModKeyCode for key input events, which also seems to fix the problem for qt5. Some additional discussion is also in tdf#103158. Regression from commit 862fdb98ca271b60a831cd5420fd16d5f9c1c747 ("tdf#143298 Qt5 send SalEvent::KeyModChange events"). Change-Id: Iafcd1db7abcdf078001ca0602ae6e374f2a169ce Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121858 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de> (cherry picked from commit 61cb81d67ebf6b342a1cdb46bf6eb25a49eb5ff4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121887 Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
2021-09-13Resolves: tdf#144457 Use weld::set_text() instead of replace_selection()Eike Rathke
... to replace entire field with modified text, instead of inserting the text at the cursor position without selection, thus effectively duplicating it (modulo modification). Fallout from commit 08101a1ab3b5d7c41488e93a2af518462286844f CommitDate: Tue Jul 30 14:04:17 2019 +0200 weld OfaLanguagesTabPage that did if (bModified) - { - // Do not use SetText(...,GetSelection()) because internally the - // reference's pointer of the selection is obtained resulting in the - // entire text being selected at the end. - Selection aSelection( rEd.GetSelection()); - rEd.SetText( aBuf.makeStringAndClear(), aSelection); - } + rEd.replace_selection(aBuf.makeStringAndClear()); replacing the workaround needed for the old toolkit with something similar from the new toolkit but behaving differently.. Change-Id: I9ff325eecd747bbecb36eb2a1150ae4472e475e8 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122000 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit db90a6cedbc261ad711ff13c4f69db65946486da) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121978 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-12tdf#144085 XUsedAreaCursor broken: the methods don't accept formatted cells ↵Noel Grandin
as used This reverts commit 61386aa03cd166473a58dbb4be0dd5e0ce82195c. tdf#79049 speed up OOXML workbook load (3) which caused this problem Change-Id: I19a9cf75e8b2d3b5f87e1bd324c42a1dbae8ee63 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121973 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk> (cherry picked from commit d8b765e6c45792bff5717cd0e0d9d42311c33461) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121977
2021-09-11Resolves: tdf#143777 Exclude recipient is always greyed outCaolán McNamara
Change-Id: I6ce94fad6ba518457665ae6d6b473cfe6f80849f Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121882 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-09-10tdf#126426: support for hyperlinks in TextParagraphContextVasily Melenchuk
Usually hyperlinks are processed by TextBodyContext, but for grouped shape we accidentaly gone into TextParagraphContext It has almost all possibilities to process txbxContent, but not hyperlinks. Additionally some hyperlink char attributes (color and underline) can expand to follow up ordinal text. Additional small hack applied to avoid this. Unfortunately this is not a final solution: such document fails roundtrip and hyperlinks are lost after saving to DOCX. Conflicts: oox/source/drawingml/textbodycontext.cxx Change-Id: Ie954f53696bd872cb1f59cb586fb55f6cd7c73bc Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121172 Tested-by: Jenkins Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121904
2021-09-10drop stray fprintfCaolán McNamara
Change-Id: Ide4df588f8642b3eaad6dab4c819ba1914f3e2ba Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121880 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-09-10xmloff: ODF export: avoid mysterious crash while sorting shapesMichael Stahl
Apparently the draw page contains a null XShape. That sounds like a bug but OTOH this sorting feature from commit 9bc6160e0acbc78be998129ea55ed7e4529959faa isn't that important so let's sweep the problem under the rug. 0 swlo.dll sw::GetZOrderLayer::operator()(com::sun::star::uno::Reference<com::sun::star::beans::XPropertySet> const &) C:\cygwin64\home\buildslave\source\libo-core\sw\source\filter\xml\zorder.hxx:37 1 mergedlo.dll xmloff::FixZOrder(com::sun::star::uno::Reference<com::sun::star::drawing::XShapes> const &,std::function<unsigned int > const &) C:\cygwin64\home\buildslave\source\libo-core\xmloff\source\draw\shapeexport.cxx:1003 2 swlo.dll SwXMLWriter::Write_(com::sun::star::uno::Reference<com::sun::star::task::XStatusIndicator> const &,rtl::OUString const &) C:\cygwin64\home\buildslave\source\libo-core\sw\source\filter\xml\wrtxml.cxx:190 https://crashreport.libreoffice.org/stats/crash_details/045adea4-c577-4164-9e69-bde5f892bd17 Change-Id: I1e67dc1c354cb14717cf9667314d6752e1b4c295 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121860 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de> (cherry picked from commit a0d28d03ff40a77a2c88704a7b2bedb68e68563f) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121878 Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-09-10tdf#120129: don't forget to update buffer size to actual lengthMike Kaganski
Otherwise extra bytes get written to the resulting string from the too long buffer. Change-Id: Iccde16b8002f214df6f86f484f288ec464c6b674 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121872 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> (cherry picked from commit 541ddf4580cac8c3f9590be26a487f5fc8e2553c) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121874 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
generated by cgit (git 2.34.1) at 2024-07-31 15:27:38 +0000