- 17 Eki, 2015 33 kayıt (commit)
-
-
Michael Meeks yazdı
Change-Id: I5c331380b872429cf1fa3908c6add8de879c9ed2 Reviewed-on: https://gerrit.libreoffice.org/18927Reviewed-by:
Eike Rathke <erack@redhat.com> Tested-by:
-
Tor Lillqvist yazdı
Change-Id: I60c7540241c41f5063736f2a3d4817371411c8dc (cherry picked from commit a799d7c9) Reviewed-on: https://gerrit.libreoffice.org/18716Tested-by:
Jenkins <ci@libreoffice.org> Reviewed-by:
David Tardon <dtardon@redhat.com> (cherry picked from commit 183129ea)
-
Winfried Donkers yazdı
provide forward compatibility for name change of FINV Reviewed-on: https://gerrit.libreoffice.org/18814Reviewed-by:
Christian Lohmaier <lohmaier+LibreOffice@googlemail.com> Tested-by:
-
Caolán McNamara yazdı
Change-Id: Ia174feec73ee676567a3632d2f88b11c176b6363 (cherry picked from commit b107353a) Reviewed-on: https://gerrit.libreoffice.org/18902Tested-by:
Miklos Vajna <vmiklos@collabora.co.uk> (cherry picked from commit f86ff1a0)
-
Stephan Bergmann yazdı
...it leads to better handling of the failure downstream, if e.g. this is an attempt to copy a file from the presets to an existing file of a previously partly generated UserInstallation, and the presets and the UserInstallation are on different drives and rename(3) fails with EXDEV. (Without this fix, removing the UserInstallation's registrymodifications.xcu caused restarting soffice to fail with "User installation could not be completed" in this case; regression introduced with 6edbcc1b "cid#1242936 handle rename failure.") Change-Id: I409698c4e2e6cc6ccd8b2be80a393c9657b5125d (cherry picked from commit d2a7abeb) Reviewed-on: https://gerrit.libreoffice.org/18898Tested-by:
-
Mark Hung yazdı
vertical layout. Change-Id: I6391e665db205545f0d660e3de826755c954f286 Reviewed-on: https://gerrit.libreoffice.org/17836Reviewed-by:
Caolán McNamara <caolanm@redhat.com> Tested-by:
-
Caolán McNamara yazdı
Change-Id: Ia8167833b54bcfc68f1a306e58a1134a38d64ef6 Reviewed-on: https://gerrit.libreoffice.org/18844Tested-by:
Cor Nouws <cor.nouws@documentfoundation.org> Reviewed-by:
Adolfo Jayme Barrientos <fitojb@ubuntu.com> (cherry picked from commit ce75966a)
-
Eike Rathke yazdı
Change-Id: Id0fec512b53aab47f9951056e8726af0cce45708 (cherry picked from commit 624ab2e5) Reviewed-on: https://gerrit.libreoffice.org/18664Reviewed-by:
Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by:
-
Mark Hung yazdı
tdf#40091 Vertical writing inferior to MS Word tdf#40335 Using Right-to-left (vertical) for text, the characters overlap each other. Cursor is placed at the half of the character. It seems that current implementation use HarfBuzz OpenType layout API hb_buffer_get_glyph_positions() to layout the text, however instead of setting text direction to HB_DIRECTION_TTB, it uses HB_DIRECTION_LTR as its direction and rotates each CJK glyph to simulate the HB_DIRECTION_TTB. However the API use font width to position each glyph no matter whether the glyph will be rotated or not. Change-Id: Ie35238dd92218f8c4463c0ae1d356ad8609c5f48 Reviewed-on: https://gerrit.libreoffice.org/17774Tested-by:
Norbert Thiebaud <nthiebaud@gmail.com> Reviewed-on: https://gerrit.libreoffice.org/18878Reviewed-by:
-
Mark Hung yazdı
Language poolitem will be inserted after user completes editing with IME, making it refer to valid range when undo. Change-Id: Id2876aa74dba6f7d134b8e2df4d9b36a8f429bb1 Reviewed-on: https://gerrit.libreoffice.org/18392Reviewed-by:
-
László Németh yazdı
AutoCorrect Emoji replacements were incomplete in Calc cells and Impress text boxes, keeping the terminating colon: :omega: -> Ω: Corrected by this patch: :omega: -> Ω Change-Id: I0d1f6f9ec9c31a7b37e0c9afaaad17dcee568dd5 Reviewed-on: https://gerrit.libreoffice.org/18849Reviewed-by:
-
Christian Lohmaier yazdı
Change-Id: I6287394ba04c7ec81ffb39229a0147e829615b07 (cherry picked from commit dfb00ac6) (cherry picked from commit e0910afb)
-
Eike Rathke yazdı
... if more than one argument is passed to PRODUCT() or SUMSQ() and matrix/array arguments are involved the functions returned wrong results. For example, wrong behavior =PRODUCT({2,3},{4,5}) gave 6 =SUMSQ({2,3},{4,5}) gave 34 Correct is =PRODUCT({2,3},{4,5}) result 120 =SUMSQ({2,3},{4,5}) result 54 Change-Id: Iec7b136a5cc678416f28b1abfe19cd0901ef22b6 (cherry picked from commit c1780ad2) Reviewed-on: https://gerrit.libreoffice.org/18864Reviewed-by:Łukasz Hryniuk <lukasz.hryniuk@wp.pl> Tested-by:
-
Cor Nouws yazdı
this partially changes behaviour introduced with commit bb00a009 but without the performace problems from the old code thanks to kendy for helping (cherry picked from commit 2fcf8923) Change-Id: I5dab81b58262f67db1c70223c612636a8b8c90ee Reviewed-on: https://gerrit.libreoffice.org/18816Reviewed-by:
-
Lionel Elie Mamane yazdı
fixes a segfault (crash) in these conditions: * the form's AfterInsert StarBasic script does a reload() of the form * the insert happens by clicking into a cell which has scripts linked to the events FocusLost and FocusGained. The case that led to discovering this involved a ListBox whose RecordSource changes when focus is gained or lost. Then there was a race condition between the cell's dispose (from the form's reload()) and the _old_ cell's FocusLost being executed. NewStyleUNOScript::invoke would then call xControl->getModel with xControl being the involved cell, leading FmXGridCell::getModel() to call m_pColumn->getModel() with m_pColumn == NULL. Change-Id: Ifb4402d37ee4faec80087ffccabe102acc016d60 Reviewed-on: https://gerrit.libreoffice.org/18669Tested-by:
-
Miklos Vajna yazdı
DocxAttributeOutput::SectionBreaks() previously only handled the text-text and text-table node transitions; implement support for table-text to avoid loosing a page break on export for the bugdoc. (View this commit with whitespace ignored to filter out the noise about SectionBreaks() now accepting non-text nodes, too.) Change-Id: Ie8a1575374a207399351635bda8c0c076ce7268d (cherry picked from commit c916152d) Reviewed-on: https://gerrit.libreoffice.org/18759Reviewed-by:
-
Julien Nabet yazdı
Just use the right order LOCAL_PUBLISHER first, ZIP_PUBLISHER second, FTP_PUBLISHER third and last one Change-Id: I23ff96785910d62bbb2c221932d0ad2a2e755b3c Reviewed-on: https://gerrit.libreoffice.org/18755Tested-by:
Julien Nabet <serval2412@yahoo.fr> (cherry picked from commit 09c3e186) Reviewed-on: https://gerrit.libreoffice.org/18758Reviewed-by:
-
Julien Nabet yazdı
Adjust some values for make room (ok with French, German and English UI) Change-Id: I7600acae64f5ae1110c07798cc950ec73a48deab Reviewed-on: https://gerrit.libreoffice.org/18787Reviewed-by:
-
Julien Nabet yazdı
Value of 60 allows to display the selected dates + let a margin if year is in 4 numbers one day. Change-Id: I3d7db50dd318897b0b7a515f9428401c53218c08 Reviewed-on: https://gerrit.libreoffice.org/18786Tested-by:
-
Eike Rathke yazdı
Change-Id: Iaadf7044acb7299ee878de2fbc08992447587d62 (cherry picked from commit f09a9163) Reviewed-on: https://gerrit.libreoffice.org/18751Tested-by:
-
Caolán McNamara yazdı
Change-Id: I5d1d18f864496ec98894703f0a22e211a9dd14a5 (cherry picked from commit a2b90cd0) Reviewed-on: https://gerrit.libreoffice.org/18747Tested-by:
-
Katarina Behrens yazdı
The fix is twofold: 1.Get URL property from the underlying text field, not from the text run -- put text field properties into rXPropSet (that's what GETA macro later queries), not into rRun 6a043e9c does s/rXPropSet/rRun/ afaics for no good reason 2. Retrieve string content from URL field early, so that the test for empty text content doesn't fire Reviewed-on: https://gerrit.libreoffice.org/18031Tested-by:
-
Caolán McNamara yazdı
this is presumably a regression from commit 8904bf48 Author: Michael Stahl <mstahl@redhat.com> Date: Tue Jul 7 23:57:07 2015 +0200 sd: dispose SdStyleSheet properly If it's disposed during document destruction it still retained pointers to SfxItemSet and SfxStyleSheetBasePool. where it was always deleted in the dtor, but now is optional in ::dispose lets set bMySet to true when pSet is created like sc does, rather try going to an unconditional delete in ::dispose Change-Id: I214675805e513f657e51c779758106194f6d14e3 (cherry picked from commit 217a2218) Reviewed-on: https://gerrit.libreoffice.org/18694Tested-by:
-
Caolán McNamara yazdı
(cherry picked from commit 7af19f45) if nSize was 1 with unicode encoding nSize would wrap around in lcl_getMaxSafeStrLen (cherry picked from commit f6368c29) rework this so we don't read the string, backup and re-read the string again (cherry picked from commit 8523e57e) Change-Id: I99f3d4a2ec760228f485d01fce856deb9c068431 Reviewed-on: https://gerrit.libreoffice.org/18690Tested-by:
-
Caolán McNamara yazdı
regression from commit 820576af Author: Noel Grandin <noel@peralex.com> Date: Thu Mar 19 13:54:12 2015 +0200 start wrapping OutputDevice in VclPtr 9,800 (568 direct, 9,232 indirect) bytes in 1 blocks are definitely lost in loss record 12,696 of 12,898 by 0xC602E72: ScopedVclPtrInstance<VirtualDevice>::ScopedVclPtrInstance<>() (vclptr.hxx:375) (cherry picked from commit 23d4d2f6) Change-Id: I356f39c339fd28a9e19a00d61b6f1bee492f2638 Reviewed-on: https://gerrit.libreoffice.org/18689Tested-by:
-
Julien Nabet yazdı
0 strlen () at ../sysdeps/x86_64/strlen.S:106 1 0x00007fffdfd69083 in SalDisplay::GetKeyNameFromKeySym (this=this@entry=0x1121b30, nKeySym=<optimised out>) at /build/libreoffice-Ke3JzN/libreoffice-5.0.1~rc2/vcl/unx/generic/app/saldisp.cxx:744 2 0x00007fffdfd696f5 in SalDisplay::GetKeyName (this=0x1121b30, nKeyCode=nKeyCode@entry=9476) at /build/libreoffice-Ke3JzN/libreoffice-5.0.1~rc2/vcl/unx/generic/app/saldisp.cxx:798 3 0x00007fffe17892a2 in GtkSalFrame::GetKeyName (this=<optimised out>, nKeyCode=<optimised out>) at /build/libreoffice-Ke3JzN/libreoffice-5.0.1~rc2/vcl/unx/gtk/window/gtksalframe.cxx:3005 4 0x00007ffff6125bbc in vcl::KeyCode::GetName (this=this@entry=0x7fffffffbfb0, pWindow=<optimised out>, pWindow@entry=0x0) at /build/libreoffice-Ke3JzN/libreoffice-5.0.1~rc2/vcl/source/window/keycod.cxx:108 Thank you Alex Bennée for having noticed this. Change-Id: I4fa89241c72144d5c76cd6175368c6878fc57a46 Reviewed-on: https://gerrit.libreoffice.org/18708Reviewed-by: -
Julien Nabet yazdı
pubAware_ adds 2 elements for each of these: LOCAL_PUBLISHER, ZIP_PUBLISHER, FTP_PUBLISHER and since ZIP_PUBLISHER is added in second positon, the right index is 2 not 4 Change-Id: I0a533fd0cefcbbfca0e95628f64fac6258019600 Reviewed-on: https://gerrit.libreoffice.org/18715Reviewed-by:
-
Maxim Monastirsky yazdı
Regression of 46666a77. The code in MenuBarManager::FillMenuManager merges the addon menu before .uno:WindowList, which doesn't exist in the start center since that commit. As a last resort, let's also check for .uno:HelpMenu (assuming that it's never placed before the Window menu). Change-Id: If45eebe4351c40d8ed69daba527844ffc02e8458 Reviewed-on: https://gerrit.libreoffice.org/18729Reviewed-by:
Jan Holesovsky <kendy@collabora.com> Tested-by:
-
Eike Rathke yazdı
... which is what Excel does, the value has to fit in a 16-bit signed integer. Change-Id: Ie172eb4eaa69b80110c8f93b1f74a6bb17ffc92f (cherry picked from commit ef725dc3) Reviewed-on: https://gerrit.libreoffice.org/18685Tested-by:
-
Caolán McNamara yazdı
Change-Id: I4ee52c76b2c1723639185bc99b85d83be5b4fdd7 (cherry picked from commit 637ad743) Reviewed-on: https://gerrit.libreoffice.org/18686Reviewed-by:
-
Rene Engelhard yazdı
Needed since 2ce903c7 Conflicts: configure.ac Change-Id: I589e782baae0bd4d5906b97c371cad587aaf5c94 (cherry picked from commit 8d5fda8e)
-
Andras Timar yazdı
... because special 'debian-menus' package from sysui module does the job We do not have these scripts in rpm packages, let's not have them in deb packages, too. Change-Id: Iae7e996c3c3a8c1fb40977436773829a16e93fee (cherry picked from commit f8f1a064)
-
Andras Timar yazdı
... tools don't find *.desktop files in /usr/local or elsewhere, and the applications will not be registered in menus. Fixed the regression from d2cc121e Change-Id: I679b98b5cd0a502553fb13025ee8b9ac4bb5507e (cherry picked from commit eb911d69)
-
- 16 Eki, 2015 7 kayıt (commit)
-
-
Jan Holesovsky yazdı
The slides to export can be tweaked via a "PagePos" parameter. Change-Id: I66f19521bd8f699710eefafb29f54036d7e604c3 (cherry picked from commit e0769daf)
-
Jan Holesovsky yazdı
Change-Id: I0de8b5229a83d2ed1ccecbd5608c94881b9baad3 (cherry picked from commit 2cb17258)
-
Miklos Vajna yazdı
(cherry picked from commit d8bc93ef118c75f48685d5f6d121ad7eabf2a80c) Conflicts: sd/source/ui/view/Outliner.cxx Change-Id: Ib6be973bf7c911907c793571b84021dc317dcc80
-
Miklos Vajna yazdı
Given that we emit selection callbacks on each logical operations (set selection, deselect all, etc) already, it's not necessary to emit callbacks in DrawSelection() when it's called from Paint(). Change-Id: I6781c71ee2f292de2a1c7d129fc49b514aca2be1 (cherry picked from commit c8bfca547d822a75ddfdd33903deae3fb6488789)
-
Miklos Vajna yazdı
Change-Id: I3f34f0315045d33ff6e498e24c0dacb0aabb0ff9 (cherry picked from commit 293dd731)
-
Miklos Vajna yazdı
Change-Id: I607b3719e0f508f9ae24db7482323847aa8e2491 (cherry picked from commit 6a35a75a)
-
Miklos Vajna yazdı
The LOK API can describe a multi-selection, so find-all can signal all matches, editeng can have a single selection only. Instead of having no selections after a find-all, select the first match, so e.g. copy works. Change-Id: I0eab2565916f0c3cce5d77279c0d927ad4b7054c (cherry picked from commit cd497698)
-
