summaryrefslogtreecommitdiff
path: root/vcl
AgeCommit message (Collapse)Author
2022-01-26avoid Xlib cairo surfaces for small virtual devices (bsc#1183308)Luboš Luňák
The (private :( ) document contains a large number of small shapes for which VclProcessor2D::RenderTransparencePrimitive2D() gets called, which results in GetBitmapEx() on the VirtualDevice. And Cairo normally needs to do a roundtrip to the XServer to fetch the content, which makes the rendering pretty slow. Forcing image-based surface for small sizes of VirtualDevice actually has better performance for the document, and the lack of possible HW acceleration generally shouldn't matter for a surface this small. Additionally drawinglayer's VirtualDevice reusing tries to reuse even a large one when a small one is needed, so a hack is needed to avoid that in this case, I couldn't find a better way. Change-Id: I0f58659ab88165a6bd915f100fc3b1d4802a0fa9 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128309 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit cf9be3417bc2be5f772c03180b7cbd248b82cad5) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128815 Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2022-01-25tdf#146941 qt: QtFilePicker::finished needs SolarMutexGuardMichael Weghorn
Change-Id: I4165c72ddd6370c8fe7faf55f9db831359f6f8a5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128923 Tested-by: Jenkins Reviewed-by: Michael Weghorn <m.weghorn@posteo.de> (cherry picked from commit a85908f507f39c88a6cf244f5f4120361ad5e2ae) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128826 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-25gtk3: "expand" icon in reference windows goes missing on entering shrunk modeCaolán McNamara
Change-Id: Ie303f82d1cc0de10ebd14248a31111a16be344f5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128814 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2022-01-25tdf#146971 changing a11y desc to replace %PRODUCTNAME has perf issuesCaolán McNamara
so leave it alone, and do the conversion just for the originally report situation as a safely backportable change with a follow up to not allow us to get into this situation in the first place Change-Id: I4f95f85791d0f937e53d7541804870b2cbf62b44 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128820 Tested-by: Jenkins Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2022-01-20tdf#141340 PDF export: fix hyperlinks on the wrong page with page num rangeMiklos Vajna
Regression from commit 01dbb38680aa39a4d3bc7afd05d44a4b2c9bc6ab (tdf#61274 sd PDF export: fix links ending up on wrong pages with hidden slides, 2020-03-10), the problem was that the sd/ code that mapped page numbers between the model and the PDF output only handled hidden slides, but not partial exports. Fix this by revisiting the decision to handle hidden slides in sd/, the filter/ code at the end does have enough information to correctly do this mapping at the end, and this way both tdf#61274 and tdf#141340 can work at the same time. (cherry picked from commit 31bbaf478783e3a2c21bd3ea94ddf39299a63d1e) Change-Id: I5679743ca67fab562e14c73e32f1a06ead8e7a31 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128647 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-16Qt notify LO of many QScreen related eventsJan-Marek Glogowski
LO just has a single display event, SalEvent::DisplayChanged, Qt has a multitude. So this will generate multiple LO notifications. I don't see any reasonable way to prevent that. Doesn't seem a problem here. Change-Id: I2f11e53765869a4ca292a1d337347e4e7470e3c2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128474 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de> (cherry picked from commit ced4cde59eb51e4364daf69520a184b55c2c54db) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128434 Tested-by: Caolán McNamara <caolanm@redhat.com> Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2022-01-11PDF export of PDF images: preserve hyperlinksMiklos Vajna
The input file has a single page, with a full-page PDF image, which contains a hyperlink. Similar to pdfcrop, we used to wrap this into a form XObject, so page-level annotations like hyperlinks were lost. Explicitly merge page-level annotations from the source page to the page that contains the PDF image to preserve those annotations. Change-Id: I96e8bc9d33440b91f3514486d6a8bd75047546b2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128108 Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Tested-by: Jenkins (cherry picked from commit 1984a5c140cc3c7c291047dacf3bddd7061d2047) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128156 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-10tdf#73523 support WinBMPv3-NT version of BMP files (found in Visio files)Andras Timar
It is a special case when there are additional 12 bytes of color bitfields after the info header. See also the useful comments at https://searchfox.org/mozilla-central/source/image/decoders/nsBMPDecoder.cpp#47 It turned out we generate this type of BMP files internally from EMF files, but in that case the bfOffBits of BITMAPFILEHEADER struct points to the pixel data. In case of Visio, bfOffBits points to the color mask. There was no better solution than preventing jumping back in the stream. Change-Id: I97f0fd4f184ea77beeab8b22fc98fdb78472d9bb Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127920 Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com> Reviewed-by: Miklos Vajna <vmiklos@collabora.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127927 Tested-by: Jenkins Reviewed-by: Andras Timar <andras.timar@collabora.com> (cherry picked from commit 3a0d539b3a8a76efc2cdac04c658645fe750bec4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128158 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-09tdf#146641 allocations attempted while hidden are discarded by gtkCaolán McNamara
Change-Id: I46288cf4c106e2763feba298f1c44dbbf6c85581 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128080 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-09Resolves: tdf#129745 don't override child help for application framesCaolán McNamara
the glue WindowInstance is created on demand, and on creation descends into any existing open dialogs and changes their help handlers. We could not descend into child dialogs, but we shouldn't change the app frame help handler anyway. Change-Id: I8d149d324f04824fbe9eec517d8c482fd96b125b Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128075 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-04Always use <poll.h>, no modern system needs <sys/poll.h>Ismael Luceno
This was standardized in IEEE 1003.1-2001, so more than 20 years old. Verified following systems using online resources: - FreeBSD 3.0 (1998) - NetBSD 1.3 (1998) - OpenBSD 2.0 (1996) - CentOS 5.0 (2007) - Debian 4.0 (2007) - IRIX 6.5.30 (2006) - SunOS 4.1.3 (1992) - SUSE 10.2 (2008) - Red Hat 5.0 (1998) So the check used for BSDs on vcl/unx/generic/dtrans/X11_selection.cxx was never correct. On GNU/Linux specifically, <poll.h> is provided since glibc 2.0 (1997). musl libc produces the following warning if the non-standard header is included: /usr/include/sys/poll.h:1:2: warning: #warning redirecting incorrect #include <sys/poll.h> to <poll.h> [-Wcpp] Change-Id: Ia8f4b9e1ee069f86abe03140c18a77d17336d09c Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127666 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de> (cherry picked from commit 4242c7b91887236375e08e2b6b24de0cecbf1626) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127586 Reviewed-by: Ismael Luceno <ismael@iodev.co.uk> Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-22tdf#131467 Qt set default position on first resizeJan-Marek Glogowski
Setting the position in Show() is too late, because LO will try to set the mouse pointer to the default button, if configured. That obviously needs the window position. And also take the menubar offset into account. Change-Id: Ia280539c082ff6f675966869fb6643a41a17d696 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127154 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de> (cherry picked from commit ca28826a087245686d7fca3ffc8ca1f03307924d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127185 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-12-21Qt use asChild for pointer positioningJan-Marek Glogowski
.. and refactor some variable names. Change-Id: I7d31a7a377f614bdd78f8d09852a2eda0cc47442 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127153 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de> (cherry picked from commit e63fe68fb1d0915b64fdf63f7fa6eed866fa3a0d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127173 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-20Windows format name is UniformResourceLocatorW for Unicode stringsMike Kaganski
See also commit 5fb9f4ffa9284c7248e2e82210506babaad4044d tdf#145964: Windows format name is FileNameW for Unicode strings TODO: replace one remaining format from CFSTR_* family that we use (see https://www.codeproject.com/Reference/1091137/Windows-Clipboard-Formats): FileGroupDescriptor -> FileGroupDescriptorW. That one needs more complex handling. Change-Id: I4d4ad83099854768cf36c7b3d89059d79c8e77f6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126213 Tested-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> (cherry picked from commit 52e1d0ca6ad38b4b4fdc77b0951ad26f0ac18ec5) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126239 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-17Related tdf#145910: relax even more denylist about OpenGLJulien Nabet
Comment this one: <!-- tdf#125516: crash on preview of slide transitions, or in slideshow, when OpenGL rendering enabled Windows 10, with Intel DCH packaged driver --> <entry os="10" vendor="intel" compare="between_inclusive_start" minVersion="26.20.100.6861" maxVersion="26.20.100.7584"><!-- tdf#125516 --> Revert these: 1) <entry os="all" vendor="intel" compare="less" version="10.18.14.4264"> <device id="all"/> </entry> commit 5621762a36483f7bc555dd0e6a294eb68100490a Author: Tor Lillqvist <tml@collabora.com> Date: Fri Jan 8 15:53:54 2016 +0200 Don't use the "marketing" version number for the Intel driver The 15.x.y.z.d number is not the real version number that our code sees. commit b1878ab683adeff6d151617fcd8f4a4530366e0e Author: Tomaž Vajngerl <tomaz.vajngerl@collabora.com> Date: Fri Nov 20 22:55:38 2015 +0100 opengl: populate blocklist 2) <entry os="8" vendor="intel" compare="equal" version="10.18.10.3308"><!-- Intel(R) HD Graphics 4000 --> <device id="0x0166"/> </entry> commit 5e416099f088a2f8a8980e08e3d5b731da0a6d9c Author: Marina Latini <marina@studiostorti.com> Date: Thu Nov 10 14:34:00 2016 +0100 Windows 8 driver blacklist Blacklisted intel driver for graphics card Intel(R) HD Graphics 4000 for Windows 8. With this card LibreOffice won't start. 3) <entry os="7" vendor="intel"> <device id="all"/> </entry> commit c9b2af045acc92c8665a8523407f530cc691d5bf Author: Tomaž Vajngerl <tomaz.vajngerl@collabora.co.uk> Date: Tue Jul 26 16:21:13 2016 +0900 tdf#101138 opengl: blacklist intel drivers for Win 7 (crash in general but nothing about slide transition, let's give it a try) 4) <entry os="all" vendor="amd" compare="less" version="15.200.1062.1004"> <!-- 150.200 --> <device id="all"/> </entry> commit b1878ab683adeff6d151617fcd8f4a4530366e0e Author: Tomaž Vajngerl <tomaz.vajngerl@collabora.com> Date: Fri Nov 20 22:55:38 2015 +0100 opengl: populate blocklist 5) <entry os="all" vendor="nvidia" compare="less" version="10.18.13.5362"> <!-- 353.62 --> <device id="all"/> </entry> commit b1878ab683adeff6d151617fcd8f4a4530366e0e Author: Tomaž Vajngerl <tomaz.vajngerl@collabora.com> Date: Fri Nov 20 22:55:38 2015 +0100 opengl: populate blocklist 6) <entry os="10" vendor="nvidia"> <!-- tdf#128441 --> <device id="0x2182"/> </entry> commit ed94101d8c399f6de2e2b9b7cd31dd6b68d269a8 Author: Julien Nabet <serval2412@yahoo.fr> Date: Sat Nov 2 18:34:06 2019 +0100 tdf#128441: blacklist nvidia on Win10 deviceid 0x2182 (crash on Writer so not Slide transition) 7) <entry os="all" vendor="microsoft" compare="less" version="6.2.0.0"> <!-- 6.2.0.0 --> <device id="all"/> </entry> commit b1878ab683adeff6d151617fcd8f4a4530366e0e Author: Tomaž Vajngerl <tomaz.vajngerl@collabora.com> Date: Fri Nov 20 22:55:38 2015 +0100 opengl: populate blocklist Change-Id: I7c2c99488e74277cbbff2f0f37937ca3b2115b72 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126489 Tested-by: Jenkins Reviewed-by: Julien Nabet <serval2412@yahoo.fr> (cherry picked from commit b8f68233b8dc5a009396141fba6e47867e70f342) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126946 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-16crashtesting: Assertion `nMin <= nMax' failed in MinMaxCaolán McNamara
MinMax(T, tools::Long, tools::Long) [T = int]: Assertion `nMin <= nMax' failed. seen since commit 374e261ad1ea8b41f5ecdd850c27fdc961c4868b Date: Sun Dec 5 11:55:58 2021 +0100 increase maximum document thumbnail size from 256 to 512 but presumably always lurked previously. This commit is supposed to change nothing, just rearrange to show that the nWidth == 1 branch (and adapted the nHeight == 1 branch too) doesn't use the data generated by the asserting block and move that block into the nWidth/nHeight != 1 branches Change-Id: I1d3284ee32c1eff738c34bc252400726dc7632b4 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126895 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 5cfb92d11de73aba88dce809cc0355a8d9c23e95) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126884
2021-12-15unset mpMenuBarWidget when it was destroyed along with its parentCaolán McNamara
Resolves: https://github.com/flathub/org.libreoffice.LibreOffice/issues/173 Change-Id: I875cf658fb86adfa389429ead059bfd7c4f08ef4 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126877 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-15tdf#145934 workaround rounding causing 'dancing characters'Caolán McNamara
Change-Id: I1fb260196beb0cc54232aa60a1191d3090fa31b0 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126870 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-11Revert "tdf#145354: Ensure displayed paper name matches displayed paper ..."Tor Lillqvist
It causes tdf#146045. This reverts commit 65081542d2dabdf17820d62abdc5a22d3734e52d Change-Id: Id980701b1a823c8bf7cfb57e0b32cf5d672c3bfa Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126373 Tested-by: Jenkins Reviewed-by: Tor Lillqvist <tml@collabora.com> (cherry picked from commit d2ca15952de167b6236139d53f6757a7c0d22d0a) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126526 Reviewed-by: Justin Luth <jluth@mail.com>
2021-12-10tdf#138999 only grab keyboard and not pointing devicesCaolán McNamara
this way scrolling in a combobox dropdown and moving the mouse outside the widget doesn't cause the scrollbar to jump to the top. No ill effect seen in a) autofilter dropdowns in calc b) color dropdown in writer main window or sidebar and clicking on combobox subwidget to launch its popup and select an entry to return to color dropdown c) submenu popups from style combobox dropdown Change-Id: I6ad445a6fab899d52907237e1f5506cce31b1f48 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126408 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 4f41fb1cba455db617d7fb68450ea1e38ac9cb0d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126430
2021-12-09Related tdf#145910: relax a bit more denylist about OpenGLJulien Nabet
tdf#99919, tdf#100243, tdf#117477 and tdf#115092 are not related to slide transition Change-Id: I41e80a6b1d025adbde7bccbedb7235d4c3c8b281 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126462 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit dd4120ce811b8530505650d481514009a441234f) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126522 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-09ofz#42082 OOM with embedded pict in ww6 documentCaolán McNamara
Change-Id: I2772e55c20d4f38d26bfe36250f4fd281d4713d6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126521 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-09tdf#145322, tdf#144378 fix printing for vertical writingMark Hung
Printing still uses ExTextOutRenderer to render text because Skia haven't yet support printing and DWriteTextRenderer can't bind the printer DC. ExTextOutRenderer uses win32 API ExtTextOutW. In order to renderer upright CJK text in vertical writing with that API, the HFONT created with CreateFontIndirectW needs a font name prefixed with '@'. OTOH, use '@' prefixed font with Skia break the vertical writing unit test. - WinSalGraphics::ImplDoSetFont: use '@' prefixed font name if the requested font is vertical and is for printing. - ExTextOutRenderer: use SetTextAlign and text metric tmDescent to adjust vertical glyphs. It's not consistent with Skia or DWriteTextRenderer, and is still incorrect in many cases. The patch is adapted from reverting commit 5686c1aca40beb9514d40c86b4a3780a8a1334ba Author: Mark Hung <marklh9@gmail.com> Date: Sun May 2 14:45:45 2021 +0800 vcl: use DWriteTextRenderer for vertical writing. Change-Id: Ib2d3df8b68cad4bebe0672c9da0a16b62aed99e6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125978 Tested-by: Jenkins Reviewed-by: Mark Hung <marklh9@gmail.com> (cherry picked from commit 6720e9569d7ab6c20616ec6b97e5d4a56948908b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126443 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-09Related: tdf#145786 cooperate between our own grabsCaolán McNamara
Change-Id: I97a1868c3f086b7f414d18d0fb4daa1c6d09846f Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126514 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-07tdf#145910: relax a bit denylist about OpenGLJulien Nabet
since we don't use OpenGL anymore for rendering, only for slide transition (and some canvas part but it doesn't work) Change-Id: Ia891915fcc10c0817a79e30fa59605a9681b6536 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126461 Reviewed-by: Julien Nabet <serval2412@yahoo.fr> (cherry picked from commit 3a61cce54277fd12570103a191c50d9b37ef3dd3) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126440 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
2021-12-07fix overflow in cairo downscaled bitmap cache (tdf#137719)Luboš Luňák
In my system, sizeof(long long) == sizeof(long) == 8, so multiplying by LONG_MAX overflows long long. Change-Id: Ieb9613ef05916ef24a64db69f698036ecaf194e2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126456 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit f8ffc971545bb54aaebd227fa841f83660dba99c) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126438
2021-12-06increase maximum document thumbnail size from 256 to 512Luboš Luňák
It's 2021. Something's telling me people now care more about document previews not being blurry than an insignificant size increase of documents. See e.g. comments #21,#25 in the HiDPI bugreport tdf#144214. It also doesn't make much sense for the thumbnail creation to try hard to make the image smooth and then downscale it too much. Change-Id: I8df778dda05cf42cd27adf8f7757097fc7650acb Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126376 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit 374e261ad1ea8b41f5ecdd850c27fdc961c4868b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126301 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2021-12-04only bilinear+mipmap for Skia/raster to-screen drawing (tdf#146024)Luboš Luňák
The code already tries to hide the cost of the high-quality bicubic scaling by caching, but there are still cases where there's too much scaling done. Since this is only drawing to screen, use only bilinear+mipmap scaling in raster mode, which should be good enough (it's what the "super" scaling VCL algorithm for BmpScaleFlag::Default does as well). Change-Id: I75c86932e097411422dc1ef5e0534059dbf11ff8 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126326 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit 0a596fbd94e1750e9d1f5cdab624c719b33c3ab4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126296
2021-12-03tdf#145604 indicate the row as the target when dropping in a visual treeCaolán McNamara
and indicate between rows when dropping in a visual list, which is what we are doing for the gtk version. a visual list case is tools, customize, menu and drag and drop the assigned commands around. a visual tree case is tools, macros, organize dialogs, use libraries to create two libraries, and move to dialog and create a dialog in one library, and drop it onto the other library to move it between libraries. Change-Id: Ia8dc944c1f411b1f072ad1c8dcc755a1e44e3b05 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126291 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-03handle Skia not drawing GPU-backend shader to a bitmap (tdf#145797)Luboš Luňák
https: //bugs.chromium.org/p/skia/issues/detail?id=12685 Change-Id: I4de51b154e270df470102c357c25bdfa430a4488 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126118 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit afdfc9bb5d56fde866882e7d87869ecafed1e3d0) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126128
2021-12-03fix Skia copyArea() not coping with coordinates outside (tdf#145811)Luboš Luňák
Apparently the call is expected to handle even copies that are partially outside of the area, e.g. window scrolling seems to do this occassionally. Change-Id: I9a06c047f00d6b5b920d61f577baa9181bdc5a2b Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126147 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> (cherry picked from commit cdebd76284204f6a34df2a01d4eaedbd540c5fe6) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126130
2021-12-03tdf#145964 related: also drop special-casing in TransferableHelper::SetStringMike Kaganski
Follow-up to commit 5fb9f4ffa9284c7248e2e82210506babaad4044d Change-Id: I46d5ea404f77ac5ff67b6ee6a42afee13274a481 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126174 Tested-by: Mike Kaganski <mike.kaganski@collabora.com> Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126181 Tested-by: Jenkins
2021-12-03tdf#145964: Windows format name is FileNameW for Unicode stringsMike Kaganski
Otherwise SIMPLE_FILE clipboard format arrives encoded in ACP, and fails for any characters not representable in ACP. Change-Id: Ice8cfd98955e3ef49682aa21b41a313786b291f1 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126131 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> (cherry picked from commit 5fb9f4ffa9284c7248e2e82210506babaad4044d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126133 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-02undo "allow vcl gen menus to have non-square images"Caolán McNamara
This reverts commit 06f36cc7ca8fc056dd8cf4d8cdbe682f9a003cef ended up not using this so don't need the extra complexity Change-Id: I64a2d6f620cc864a75523be6612c5cf2086f7a00 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126211 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-02Related: tdf#145901 use x11 popover replacement for all non-menubutton casesCaolán McNamara
replacement from inside real popover under x11 as in autofilter case doesn't work. Change-Id: Ie4ba95645a8094732b9bad9829426e77375d63ad Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126208 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-01tdf#145952 don't crash with LO_COLLECT_UIINFO set in autofilterCaolán McNamara
Change-Id: I27c474c522554c825c0296cdf711d481d22fd024 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126129 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-01gtk[3|4] calculate row height closer to what gtk appears to doCaolán McNamara
we're getting slightly more lines than requested visible in f.e. the calc autofilter drop down Change-Id: I9ca16eb5ec6848499997a6fceec1d0ead693efce Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126104 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-30gtk[3|4] allow anything that would enable positioning onscreenCaolán McNamara
Change-Id: I68a96c4ae860c705108f4aaaf23e4bd05badf849 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126127 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-30tdf#145873 Fix FILEOPEN: EMF file not displayed #2Hossein
The previous patch 149bd802623d6d83ec9aa04514b938494f7f01f6 only partially fixed the problem, and the size of the graphical objects was wrongly calculated. This patch fixes this problem. The fix can be tested with: make CPPUNIT_TEST_NAME="testTdf145873" -sr \ CppunitTest_vcl_pdfexport This test exports the PPTX to PDF and checks the size of the objects. The previous test for this issue which exported to SVG is removed, as it could not test the current situation. Change-Id: I97bed99811eaf8236ded84ffe0c1a7d3d1b5bad5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125924 Tested-by: Jenkins Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl> (cherry picked from commit 20d5e320801606856c1d05c4f34d9fa82371aac8) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126102 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-11-29gtk3: set anchor_hints for gdk_window_move_to_rectCaolán McNamara
Change-Id: Id7d79e0b505cfaadf1dbae770b69eea46b43157c Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125988 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-27Related: tdf#144410 we don't want a gap in a multicol spanning separatorCaolán McNamara
we would need another column to show a drop down indicator, and we don't want the separators to have a visible gap in that case Change-Id: Ib45b4cda41a09b631f3ea4d4427a8073a9e243d6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125900 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-26-Werror,-Wreorder-ctorStephan Bergmann
Change-Id: Ie0a1036f57bc1449dda324b99fd83170969958a5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125856 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit c1d0b2ec10ee10127409a6e085b26fdc86d426ad) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125838
2021-11-25Resolves: tdf#145786 get the correct bounds with window scaling enabledCaolán McNamara
the per-window/app scaling as opposed to the global desktop one which wasn't affected. the menubutton thought it saw the mouse release happen outside the window it popped up when the combobox inside it was clicked, so popped down the popup from underneath the combobox. Change-Id: Iace9538073bb2380443d87600a872e88885934d3 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125807 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-25vcl: pInputElement is never nullptr in PDFObjectCopier::copyRecursively()Miklos Vajna
Change it to a reference to make that more obvious. Change-Id: Ie1da9982ee2d54a5d68d393a04643d1684521ef5 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125786 Tested-by: Jenkins Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
2021-11-24gtk3: I don't think it makes sense to claim a crossing event is consumedCaolán McNamara
if it does, then we should use the return of m_aMouseMotionHdl.Call() Change-Id: I9f04fcba1ef1d9768c3ce3b042884110a63f26a4 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125781 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-24Resolves: tdf#145864 we should only return true if we consume the motion eventCaolán McNamara
and false if we want it to continue to be passed on to other handlers Change-Id: Ie237f890e3d065949dcfedb08b9f6c804643b4b6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125780 Tested-by: Caolán McNamara <caolanm@redhat.com> Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-24gtk[3|4] don't need the ifdefs if only using the dlsymed functionsCaolán McNamara
Change-Id: I2c3b7a19f047d341467f9c66ef93e3185330854d Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125744 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-11-24Rename o3tl::underlyingEnumValue to o3tl::to_underlying to match C++23Mike Kaganski
Change-Id: I1597a8e77199445bf377dbe54adc3134bb04fd51 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125748 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
2021-11-24tdf#84283: vcl_pdfexport: Add unittestXisco Fauli
Change-Id: I56f50dcd93ef196d2261be924e004b9981d405dc Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125741 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-11-23CppunitTest_vcl_pdfexport: factor out common codeXisco Fauli
Change-Id: I4e25906f2b63b36d31e08282ec72921fe10f08b4 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125738 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
generated by cgit (git 2.34.1) at 2025-03-13 03:13:10 +0000