- 04 May, 2019 1 kayıt (commit)
-
-
Katarina Behrens yazdı
An overloaded drawPolyPolygon for zero transparency case used to exist and it used applyColor to fill cairo path. It was removed by commit 7034311d and the new transparency-aware drawPolyPolygon missed cairo_set_operator call that is present in applyColor. This works OK most of the time but breaks sometimes when no transparency (~no antialiasing) is involved. To fix that add transparency argument to applyColor and use it where applicable Change-Id: Ib1b077e38e7f5d30778434d45be67284407a7d16 Reviewed-on: https://gerrit.libreoffice.org/71759 Tested-by: Jenkins Reviewed-by:
Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
-
- 03 May, 2019 12 kayıt (commit)
-
-
Tomaž Vajngerl yazdı
Change-Id: I7625adbf365aa908c072ca42060e926569629044 Reviewed-on: https://gerrit.libreoffice.org/71279 Tested-by: Jenkins Reviewed-by:
Tomaž Vajngerl <quikee@gmail.com> (cherry picked from commit 43f8ea21) Reviewed-on: https://gerrit.libreoffice.org/71574Reviewed-by:
Caolán McNamara <caolanm@redhat.com> Tested-by:
-
Caolán McNamara yazdı
Change-Id: I72a0deaab3d86ac0772092ab5f60bcace7632847 Reviewed-on: https://gerrit.libreoffice.org/71732 Tested-by: Jenkins Reviewed-by:
-
Caolán McNamara yazdı
https://github.com/anholt/libepoxy/issues/180 sounds very similar and 1.5.3 apparently fixes that Change-Id: I009f5bc82f9e8326a7028ed29d86733cce649d15 Reviewed-on: https://gerrit.libreoffice.org/71750Reviewed-by:
Michael Stahl <Michael.Stahl@cib.de> Tested-by: Jenkins
-
Julien Nabet yazdı
The crash affects Windows, not Linux where an error popup appears instead (MacOs?) It seems the fact the url was empty triggers an exception in ucbhelper::Content ctr So let's put the lines in the try catch and log the exception. See comments of bugtracker for slight more details. Change-Id: I60e55be488632ff6e0106fde832a818c6431f58e Reviewed-on: https://gerrit.libreoffice.org/71321 Tested-by: Jenkins Reviewed-by:
-
Eike Rathke yazdı
It has been like that since the initial import but is utter nonsense. It might had (doubtfully) served some early StarBasic compatibility quirk, but is wrong and was implementation defined buggy behaviour. The Option VBASupport 1 even explicitly disabled it. In future we may want to implement the VBA Nothing value for an SbxEMPTY at least for boolean operators, but this for Calc user defined macro functions might even need a distinguished SbxEMPTYCELL or such. Or an explicit SbxNOTHING. Change-Id: I28919d982d0e60b9b840a12271dc717effa59662 Reviewed-on: https://gerrit.libreoffice.org/71701Reviewed-by:
Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 15c39bb2) Reviewed-on: https://gerrit.libreoffice.org/71705Reviewed-by:
Mike Kaganski <mike.kaganski@collabora.com>
-
Michael Stahl yazdı
Fixes CVE-2019-7317. Change-Id: I3374f5cbd6552e2c1569d63ee680d0c1d9389621 Reviewed-on: https://gerrit.libreoffice.org/71663 Tested-by: Jenkins Reviewed-by:
-
Caolán McNamara yazdı
Change-Id: I6e41318a92f02e3cd7fde5c52272582345362533 clamp to target type bounds Change-Id: I8d3f7653b7e9b64a2f433b4ebfb8a0fef1522e93 Reviewed-on: https://gerrit.libreoffice.org/71637 Tested-by: Jenkins Reviewed-by:
Xisco Faulí <xiscofauli@libreoffice.org>
-
Michael Stahl yazdı
The "Delete spaces and tabs at end and start of line" AutoFormat option was deleting the wrong characters: on the 2nd line, it deletes the spaces as it should, but on subsequent lines the selection is off by the number of characters that have already been removed. The range indexes are now integers, so not automatically corrected on deletions like the SwIndex they were before; in this case, where only plain characters are deleted, it should suffice to iterate the ranges in reverse order to avoid the problem. (regression from 180e5f51) Change-Id: Ie8e08c00803d94e700cd5b64f29edb0b841eda2b Reviewed-on: https://gerrit.libreoffice.org/71686 Tested-by: Jenkins Reviewed-by:
Thorsten Behrens <Thorsten.Behrens@CIB.de>
-
Michael Stahl yazdı
Avoid the problem by not moving the drawing object to the invisible layer, which isn't necessary because it will be inserted immediately anyway. (regression from 214e6ed5) Change-Id: I9936de990d1db56fa4c00c37d05b4ed2923790ff Reviewed-on: https://gerrit.libreoffice.org/71685 Tested-by: Jenkins Reviewed-by:
-
Michael Stahl yazdı
This failed several times recently on Mac in Jenkins. - Expected: 14.7945205479452 - Actual : 14.8 - Delta : 0.001 Change-Id: I966b8085bf4120afb0a09e50f85ada4e5d8d488c Reviewed-on: https://gerrit.libreoffice.org/71725 Tested-by: Jenkins Reviewed-by:
-
Stephan Bergmann yazdı
...by storing the temporary HTML document in a location that can be accessed by the browser running outside the Flatpak sandbox. This reuses and extends the mechanism already in place for the new HTML-based help in Flatpak mode (see 72b936d7 "Enable --help=html for flatpak"). This fixes <https://github.com/flathub/org.libreoffice.LibreOffice/issues/85> "“Preview in Web Browser” does not work in Flatpak version". Change-Id: I5f73fd89139ffe6b8ab0dc501154b4f054a0ae5c Reviewed-on: https://gerrit.libreoffice.org/71570 Tested-by: Jenkins Reviewed-by:
Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit c61e7b99) Reviewed-on: https://gerrit.libreoffice.org/71717Reviewed-by:
-
Christian Lohmaier yazdı
* Update translations from branch 'libreoffice-6-2' - update translations for 6.2.4 rc1 and force-fix errors using pocheck Change-Id: I9d1bb6d9641ee21cced75af9801c4ba7a68c1983
-
- 02 May, 2019 1 kayıt (commit)
-
-
Stephan Bergmann yazdı
...as flagged with -Werror=deprecated by trunk GCC in -std=c++2a mode. But C++17 forbids "this" in combination with a capture-default of "=", so in such cases list any entities explicitly that shall be captured by value. Change-Id: Ic228cd20682277a0f93e3e75798356d0fbbb80d4 Reviewed-on: https://gerrit.libreoffice.org/65241 Tested-by: Jenkins Reviewed-by:
-
- 01 May, 2019 1 kayıt (commit)
-
-
Caolán McNamara yazdı
so it can shrink to its optimal size Change-Id: Ia83b8b29a56e0e232956770a3a3374035db2c6ef Reviewed-on: https://gerrit.libreoffice.org/71577 Tested-by: Jenkins Reviewed-by:
Adolfo Jayme Barrientos <fitojb@ubuntu.com> Tested-by:
-
- 30 Nis, 2019 7 kayıt (commit)
-
-
Katarina Behrens yazdı
thus we reimplement it for Qt5DropTarget. This is qt5 remix of tdf#118302 (in Calc drop into the same tab should cancel DnD, instead of causing data loss) Change-Id: Ib37ea5a018133779e85e8e131d81bb6cee7d9206 Reviewed-on: https://gerrit.libreoffice.org/71531 Tested-by: Jenkins Reviewed-by:
Katarina Behrens <Katarina.Behrens@cib.de> (cherry picked from commit cc6c1798) Reviewed-on: https://gerrit.libreoffice.org/71569Reviewed-by:
-
Katarina Behrens yazdı
some hard to reliably reproduce crashes when drag'n'dropping slides in slide sorter in Impress can be tracked down to null drop target. Not every SalFrame is registered as drop target, so let's accept drops (QWidget::setAcceptDrops) only for those frames that are. Change-Id: I01f006d619209c558e8d9976116daad65f51d7d9 Reviewed-on: https://gerrit.libreoffice.org/71533 Tested-by: Jenkins Reviewed-by:
-
Tomaž Vajngerl yazdı
Previously faux bold was drawn with setting a dictionary key kCTStrokeWidthAttributeName and a value of the stroke width. This stopped working because we don't use attributedString for drawing anymore but we always use CTFontDrawGlyphs and draw the glyphs, which doesn't have a way to set this attributes. It took me a while to find a workaround (thanks to Apple's great documentation), where we switch text drawing to fill and stroke, and after that we can modify the stroke by changing the line width. Change-Id: I594fcb9c796a9c04ea8234a9938ca2f719706199 Reviewed-on: https://gerrit.libreoffice.org/71012 Tested-by: Jenkins Reviewed-by:
Andras Timar <andras.timar@collabora.com>
-
Tomaž Vajngerl yazdı
On macOS it is possible to copy from a PDF a subset of a document, which will be transported as a new PDF document containing the subset. LibreOffice didn't support PDF as a valid clipboard format and previously it also didn't support showing PDFs inside the document, so in such cases it copy-pasted a low resolution bitmap. The result wasn't good. As we are now able to display PDF documents as Graphic in LO, we can also support this use-case. This adds support for the PDF documents as a clipboard format in general and to the macOS backend. This commit only adds support for Writer. Reviewed-on: https://gerrit.libreoffice.org/71364 Tested-by: Jenkins Reviewed-by:
Miklos Vajna <vmiklos@collabora.com>
-
Caolán McNamara yazdı
FillInfo is called more often than Initialize, so we've ended up with multiple AcceptChgDat strings Change-Id: I99e9398757d63e2c6315ec9c3101910d5978b13b Reviewed-on: https://gerrit.libreoffice.org/71120 Tested-by: Jenkins Reviewed-by:
-
Caolán McNamara yazdı
to make it work under gtk3 Change-Id: I4c895181af233a16292ce16b194f0a60994446fb Reviewed-on: https://gerrit.libreoffice.org/71479 Tested-by: Jenkins Reviewed-by:
-
Eike Rathke yazdı
This solves the cases of Oromo, Kinyarwanda and Sinhala. Kabyle is a problem on its own and apparently not easily solvable, see attempts in https://gerrit.libreoffice.org/71247 Change-Id: Id001aeba07a95f283b4cd6057c1f6e8088de940a Reviewed-on: https://gerrit.libreoffice.org/71535Reviewed-by:
-
- 29 Nis, 2019 9 kayıt (commit)
-
-
Caolán McNamara yazdı
Change-Id: I422131ab203eba62ed5cf6fb2e19e23325b43f6e Reviewed-on: https://gerrit.libreoffice.org/71434 Tested-by: Jenkins Reviewed-by:
-
Caolán McNamara yazdı
Change-Id: I015022c4dd68f152ef5e56947031ccf4bfaf2b1c Reviewed-on: https://gerrit.libreoffice.org/71517 Tested-by: Jenkins Reviewed-by:
-
Stephan Bergmann yazdı
...from Content::getPropertyValues. ca030879 "tdf#121337: Fail on GIO error in GIO UCP getPropertyValue" had made Content::getPropertyValues fail for every getGFileInfo failure. However, when saving to a not-yet exisiting file, SfxMedium::Transfer_Impl (sfx2/source/doc/docfile.cxx) requests the properties "Title" and "ObjectId" from the Content representing the not-yet existing file, and apparently expects those requests not to fail. So restructure Content::getPropertyValues to only call getGFileInfo when actually needed (that covers not failing for the unknown- anyway "ObjecdtId" property), and handle "Title" specially by not failing for a non-existing file. (The documentation at offapi/com/sun/star/ucb/Content.idl says for the "getPropertyValues" command that: "The execution will not be aborted, if there are properties requested, that are unknown to the content." But that leaves it somewhat unclear whether failure to obtain a known property should be propagated. It apparently should be in the context of tfd#121337 fixed previously, but apparently not for "Title" here.) Change-Id: I12a9a5bd93d661922ea3b7b19a84a7e73e7e4b50 Reviewed-on: https://gerrit.libreoffice.org/71515 Tested-by: Jenkins Reviewed-by:
-
Katarina Behrens yazdı
since 7034311d affine transformation is passed down to drawPolyPolygon and drawPolyLine functions. The transformation was correctly applied to cairo context, but not the extents of damage so those had old, untransformed coordinates and were therefore not redrawn Change-Id: I61ce005ef8770ce5c4560a9d953cd92f440043be Reviewed-on: https://gerrit.libreoffice.org/71383Reviewed-by:
-
Caolán McNamara yazdı
Change-Id: I05ec394ee83f0422ecccf3ae3377818af6feb1d1 Reviewed-on: https://gerrit.libreoffice.org/71398 Tested-by: Jenkins Reviewed-by:
-
Caolán McNamara yazdı
Change-Id: I1f46c6043b3263ef983991ffa4aa0418006d55a7 Reviewed-on: https://gerrit.libreoffice.org/71292 Tested-by: Jenkins Reviewed-by:
-
Miklos Vajna yazdı
Commit a51b7a1c (tdf#103831, tdf#100986: Force using GDI when needed, 2017-03-03) noted that the DirectWrite text renderer doesn't support stretched text, add support for this now by setting a DirectWrite transform matrix that only does horizontal scaling. With this, tdf#113076 is kept fixed, but at the same time manually stretched text keeps working. Previously the glyphs of the text had the correct size and position, but the glyphs themselves where not streched, but simply aligned to the left. (cherry picked from commit 51bcbfc8) Change-Id: I8fe8e74d3edc0d71ed2f16fcce66c6f5009ed264 Reviewed-on: https://gerrit.libreoffice.org/71289 Tested-by: Jenkins Reviewed-by:
Luboš Luňák <l.lunak@collabora.com>
-
Caolán McNamara yazdı
since... commit af84fc9d Date: Tue Apr 23 15:48:41 2019 +0200 lazy image loading shouldn't read the entire .xls file (tdf#124828) nLength is just an unchecked value in the dff stream, it might not be sane so limit it to the max len of the stream Change-Id: Ia8a2830478952afe1317b5cd795f35059d9b380a Reviewed-on: https://gerrit.libreoffice.org/71413 Tested-by: Jenkins Reviewed-by:
-
Miklos Vajna yazdı
... right edge in presentation mode. This seems to be a problem since commit a51b7a1c (tdf#103831, tdf#100986: Force using GDI when needed, 2017-03-03), which blacklisted DirectWrite in the OpenGL case for both vertical text and horizontally stretched text. I tested that the overlapping characters problem from tdf#103831 is not re-introduced, and switching to DW (in the horizontal stretching && OpenGL case) fixes the unwanted clipping of glyphs for the tdf#113076 bugdoc. This also simplifies the situation, since the general rule since commit d436065b (tdf#103725: Default to GDI even with the new layout engine, 2016-11-15) is that non-OpenGL defaults to GDI rendering and OpenGL defaults to DirectWrite rendering. (cherry picked from commit 4657335f) Change-Id: I44d035ab0890614be393599a5492e42e09231d86 Reviewed-on: https://gerrit.libreoffice.org/71288 Tested-by: Jenkins Reviewed-by:
-
- 28 Nis, 2019 4 kayıt (commit)
-
-
Michael Weghorn yazdı
The separate filepicker executable 'lo_kde5filepicker' used by the gtk3_kde5 VCL plugin needs the KF5 libraries (in particular the 'KFileWidget' class) in order to provide custom controls in the Plasma-native QFileDialog. The gtk3 side ('lo_kde5filepicker') itself doesn't use any functionality from KF5 libraries itself and therefore didn't link against those. When determining which VCL plugin to use on LibreOffice start, a 'dlopen()' is used on the libraries of the potential VCL plugins, which fails if linked libraries are not available, so unsuitable VCL plugin are skipped. This mechanism didn't work for the gtk3_kde5 VCL plugin if KF5 libraries were not installed, since the 'dlopen()' on 'libvclplug_gtk3_kde5lo.so' would succeed, but starting the 'lo_kde5filepicker' process would later fail at runtime due to the missing libraries. To avoid this, make 'libvclplug_gtk3_kde5lo.so' link against KF5 libraries as well to prevent gtk3_kde5 from being used unless all required libraries are actually present. Add a dummy KFileWidget use in Gtk3KDE5FilePicker to prevent smart linkers from not linking unused libraries despite linker flags. While this is a hack, it seems to be acceptable, in particular since gtk3_kde5 is more meant as a temporary solution supposed to be dropped at a later point in time in favor of the pure kde5 VCL plugin. Reviewed-on: https://gerrit.libreoffice.org/71237Reviewed-by: -
Vasily Melenchuk yazdı
It is expected that <text:text-input> will be converted to {FORMTEXT} MS field, but practical implementation was missing. Change-Id: Ifd17a523e253555a240f853bea5f21572d90d853 Reviewed-on: https://gerrit.libreoffice.org/70786Reviewed-by: -
Caolán McNamara yazdı
so use toggle and ignore toggle offs Change-Id: Id6a9f3a7a9cf02d62ede00e4c246629406f9a64b Reviewed-on: https://gerrit.libreoffice.org/71482 Tested-by: Jenkins Reviewed-by:
-
Michael Weghorn yazdı
Do the same for the gtk3_kde5 VCL plugin as commit 2c69c93f does for the kde5 one. Change-Id: I6fc91d1dbe133bd905c033c9fce7d48c0b6464a0 Reviewed-on: https://gerrit.libreoffice.org/71384 Tested-by: Jenkins Reviewed-by:
Michael Weghorn <m.weghorn@posteo.de> (cherry picked from commit c08f23b1) Reviewed-on: https://gerrit.libreoffice.org/71410Reviewed-by:
-
- 27 Nis, 2019 5 kayıt (commit)
-
-
Caolán McNamara yazdı
spec says "An unsigned integer that MUST be 0xFFFFFFFF" for version so at least insist on that Change-Id: I8c11659d86fe122b1c8f61b1bf99e7b4b53ff306 Reviewed-on: https://gerrit.libreoffice.org/71236 Tested-by: Jenkins Reviewed-by:
-
Mike Kaganski yazdı
See thread starting at https://lists.freedesktop.org/archives/libreoffice/2018-December/081589.html Regression from commit 7263d223. That change has made unit tests DPI-aware; and then some tests started failing on systems with resolutions other than 96 DPI. It has been suggested that the proper fix would be to do for Windows what commit ada20402 did for macOS. Another approach would be to fix all the tests to be DPI-aware. I cannot do the first mentioned fix; so I have fixed testFDO74215 test in sw_ooxmlexport4; and added DPI checks to the other failing tests in chart2_xshape and sc_subsequent_filters_test to skip testing when using non-default DPI. This is not ideal, of course, and conditionally skipped tests need to be re-enabled unconditionally once a proper fix arrives. Change-Id: I5c92cfe93ae65f53a8a180fcaec49231df377b8a Reviewed-on: https://gerrit.libreoffice.org/65595Reviewed-by:
-
Caolán McNamara yazdı
Change-Id: Ifbad7984223b8d18eb353ec773705b902cab7176 Reviewed-on: https://gerrit.libreoffice.org/71368 Tested-by: Jenkins Reviewed-by:
-
Michael Stahl yazdı
Revert "tdf#123313 sw: workaround Undo problem with ToX Update" This reverts commit 60ea01af - it was quite simply a bad idea, and only pushed due to lack of time... Much better to refactor things so that any new CrossRefHeadingBookmark are inserted (with SwUndoInsBoomark) *before* nodes are inserted into the ToX. However! This patch does not fix the problem yet; that needs some more significant changes. Change-Id: If22e43b6a2112e5f53d0186bb7bed3ee007fe727 Reviewed-on: https://gerrit.libreoffice.org/71322Reviewed-by:
-
andreas kainz yazdı
Change-Id: Ibe76cc8950e26f0b78425f91faae2f44b5474ea7 Reviewed-on: https://gerrit.libreoffice.org/71264 Tested-by: Jenkins Reviewed-by:
andreas_kainz <kainz.a@gmail.com> (cherry picked from commit 5e4c1e6b) Reviewed-on: https://gerrit.libreoffice.org/71409Reviewed-by:
-
