Age | Commit message (Collapse) | Author |
|
Main problem is, that Qt::Popup grabs the focus and therefore
generates a focus-out event for the combobox, which then closes
the just shown popup window. The grab happens inside QWidget
private code and there is no way around it. But instead of
"faking" Qt::Tooltip, this uses Qt::Widget with additional flags.
Regression from commit 7e6fee830116823b9cd8e46d6962df4ea2bc1ea6
("Qt5 fix Qt::Popup window handling").
Change-Id: Ia1f8e33d98f7ec36cf1ebc350886121dfaadd658
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119691
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
(cherry picked from commit 9dcf5816c90e9819861332f11e014ef7b78e2fe7)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119621
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
Change-Id: I21fa71f47313f4edf38b2abc6e56617b519c2610
|
|
This reverts commit e35fc16969c0d0a55538733d76fc4bb03bf137b0.
Reason for revert: tdf#143456 shows there are layout problems with this compatibility option
Change-Id: I4f03b9a0cbc6de1d3dbf79681f7caa203a1ce59c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119451
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
Same as for OOXML, they are disallowed in ODFF as well.
Change-Id: Iff78a901838092a3ee32221fdaadd28ee50f7646
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119577
Tested-by: Jenkins
Reviewed-by: Eike Rathke <erack@redhat.com>
(cherry picked from commit 6eb9fde88af245a00247fca8a592ed08ee4b8bbe)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119449
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
* Update dictionaries from branch 'libreoffice-7-2'
to 447a883184eaa8643fd6ae9a85ca64ef1d912d35
- Czech Hunspell: add several word forms
Change-Id: I227d4fe75539691c75323ffcc822545081ced9ae
Reviewed-on: https://gerrit.libreoffice.org/c/dictionaries/+/119643
Tested-by: Stanislav Horáček <stanislav.horacek@gmail.com>
Reviewed-by: Stanislav Horáček <stanislav.horacek@gmail.com>
|
|
* Update dictionaries from branch 'libreoffice-7-2'
to b6f494a9946de2899acf22c843ef7386d09a5cf4
- tdf#143489: Update Galician dictionaries to version 20.08
Change-Id: I95b0064894a10a47e638568f85ddf5eb97ac345e
Reviewed-on: https://gerrit.libreoffice.org/c/dictionaries/+/119372
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>
(cherry picked from commit c292071ca8f8dece2bf5655038cdfd4899a09fa3)
Reviewed-on: https://gerrit.libreoffice.org/c/dictionaries/+/119614
Tested-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
* Update translations from branch 'libreoffice-7-2'
to 325c25d7264976f4272460ae6852f2743e62bf02
- update translations for 7.2.0 rc2/master
and force-fix errors using pocheck
Change-Id: I7b7bd01e640b88e6e81140686a68f72592e5a25d
(cherry picked from commit a2b54ae27fac12ee3aee478872632c377e1843f9)
|
|
The current code tries to on-demand create content in
form of a SwBodyFrame at a SwPageFrame. That
unexpectedly and unfortunately leads - when done
during a Table re-layout action - to an endless
layout loop.
I tried quite some solutions, but the simplest and
safest is to just supress AssertFlyPages in that
specialized TableLayout. Other solutons may break
MailMergse, so CppunitTest_sw_mailmerge needs to
be checked with any fix in this area.
For more details, see tdf#task comnments.
Change-Id: I02de9ee72848061497678936aa1d0c492694bfc2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119324
Tested-by: Jenkins
Reviewed-by: Armin Le Grand <Armin.Le.Grand@me.com>
(cherry picked from commit 0bac5c7e7d71658c5056c4bf0b71fbfb51b92ca0)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119266
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
This reverts commit 59c5732875c1bea619ed461a378906a2dce12b48.
Reason for revert: tdf#143456 shows there are layout problems with this compatibility option
Change-Id: I73cbbb73f3110bc93840b8f0de0f2fab69250ebc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119443
Reviewed-by: László Németh <nemeth@numbertext.org>
Tested-by: Jenkins
|
|
LibreOffice is only using drive.file scope, so can only see files it
owns/that were created by LibreOffice.
In addition, also store the refresh token in LO's password-store if the
user enabled persistent storage, removing the need to to the
copy'n'paste dance to grant access each time LO is launched.
related tdf#115643 also store the refresh token for onedrive
consolidate the fallback-auth provides for onedrive/gdrive into one,
they are all the same login in browser, then copy code method that
ultimately should be changed to having LO listen on local port for the
code
Change-Id: I97e3843682c302d2884e35ece6e72bc3a07e2539
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119572
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
(cherry picked from commit 73041de9563c9a973d1b5394c6e5520a7d799980)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119446
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
If LO is using list format strings (this is default behavior since
aa5c6d12) it was not able to show just numbering without all formatting,
as it used in some fields.
Change-Id: Ib4695b8e1c2d7a451522c7e04af2216d16aceefe
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119309
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
Tested-by: Jenkins
(cherry picked from commit d44730148a95933f4a45a70241cb6d1d0546f626)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119440
|
|
which bigger than 2^15 twips (> ~57.8 cm).
Regression from commit 4d1621136c464b462a598571ecdcfe2ae119d8c7
"Fix ignoring large twips values like MSO does (cp#1000087)".
Note: Writer could export page size up to 600 cm × 600 cm,
and now it can import it back as well. It can even import
page sizes in kilometers, and display that, but there will
be other issues about its display. MS Word couldn't save
page size over 2^15, but can import, and display them till
up to 2^16 (that is about 116 cm).
Co-authored-by: Tibor Nagy (NISZ)
Change-Id: Icc1bd708d53036768a7be94cb1ebe18c1c4c8487
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118940
Tested-by: László Németh <nemeth@numbertext.org>
Reviewed-by: László Németh <nemeth@numbertext.org>
(cherry picked from commit d5d9795968771651c57c661d95b6992c3c778eca)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119441
Tested-by: Jenkins
|
|
First, a document is loaded, then about 4 pages of content are pasted into
it, in about 6 paste operations.
In an idle layout action, a page's worth of content moves forward; the
text frame at the bottom was split previously, and is joined.
The follow has the same upper as the master text frame, so this causes
the upper section frame to shrink:
SwRect::operator=(const SwRect & rRect) Zeile 256
SwFrameAreaDefinition::FrameAreaWriteAccess::~FrameAreaWriteAccess() Zeile 112
SwSectionFrame::Shrink_(__int64 nDist, bool bTst) Zeile 2327
SwFrame::Shrink(__int64 nDist, bool bTst, bool bInfo) Zeile 1562
SwContentFrame::Cut() Zeile 1339
SwTextFrame::JoinFrame() Zeile 677
SwContentFrame::MakeAll(OutputDevice * __formal) Zeile 1310
The ToMaximize(false) check doesn't help because the section frame
doesn't have a follow (or footnotes).
Then the text frame is formatted, resulting in one line but no height
for it, and a split after the one line.
The split causes a follow section frame to be created, calling
SwSectionFrame::SimpleFormat() on the master's upper:
SwFrameAreaDefinition::FramePrintAreaWriteAccess::~FramePrintAreaWriteAccess() Zeile 120
SwFrame::SetTopBottomMargins(__int64 nTop, __int64 nBot) Zeile 175
SwRectFnSet::SetYMargins(SwFrame & rFrame, __int64 nTop, __int64 nBottom) Zeile 1405
SwSectionFrame::SimpleFormat() Zeile 1192
SwSectionFrame::SwSectionFrame(SwSectionFrame & rSect, bool bMaster) Zeile 110
SwFrame::GetNextSctLeaf(MakePageType eMakePage) Zeile 1781
SwFrame::GetLeaf(MakePageType eMakePage, bool bFwd) Zeile 879
SwFlowFrame::MoveFwd(bool bMakePage, bool bPageBreak, bool bMoveAlways) Zeile 1977
SwContentFrame::MakeAll(OutputDevice * __formal) Zeile 1349
SwFrame::PrepareMake(OutputDevice * pRenderContext) Zeile 286
SwFrame::Calc(OutputDevice * pRenderContext) Zeile 1794
SwTextFrame::CalcFollow(o3tl::strong_int<long,Tag_TextFrameIndex> nTextOfst) Zeile 281
SwTextFrame::AdjustFollow_(SwTextFormatter & rLine, o3tl::strong_int<long,Tag_TextFrameIndex> nOffset, o3tl::strong_int<long,Tag_TextFrameIndex> nEnd, const unsigned char nMode) Zeile 608
Now the upper has space but at this point the master text frame is valid
and not invalidated, so it never grows into the available space.
There is a check to format again in case additional space is available
in SwTextFrame::CalcFollow() but peculiarly it's disabled if the upper
is a section frame since initial CVS import.
Removing this check appears to fix the problem.
Change-Id: Ifad545f7e79675af6e33d68c7fcdbc82bd4f8f57
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119419
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 44cc59db6f4f4f8b2ce5c993a31b5a019a8d7e97)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119434
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
|
|
Accept or reject of tracked drag & drop table (row)
move weren't handled, resulting an extra empty table
(rows) instead of removing the deleted or inserted
table (rows).
Note: tables moved by drag & drop with track changes use
w:moveFromRangeStart/End and w:moveToRangeStart/End
bookmark-like elements to sign the boundary of the
moved text. Now moveFrom/moveTo import is handled by
tracked table (row) deletion and insertion.
Change-Id: Ie382645fef28f57b30d3504fb28ac08489f705c0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119406
Tested-by: Jenkins
Reviewed-by: László Németh <nemeth@numbertext.org>
(cherry picked from commit 7f3c0bbc174b9b0316991c174ca3b407c0e3d141)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119436
|
|
Change-Id: I9431221aadf97739bb197871f25fa151ef4c391c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119294
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Jenkins
(cherry picked from commit 0cda081c9aa3b3dcb363f97bac60c845ce9a13e0)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119255
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
* Fixes CVE-2020-8284 CVE-2021-22924
* Also fixes these which don't look relevant to LO:
CVE-2020-8231
CVE-2020-8285 CVE-2020-8286
CVE-2021-22876 CVE-2021-22890
CVE-2021-22897 CVE-2021-22898 CVE-2021-22901
CVE-2021-22922 CVE-2021-22923 CVE-2021-22925 CVE-2021-22926
* disable some new protocols and dependencies
* remove curl-ios.patch.1 as the code no longer exists upstream
Change-Id: I12d5f87f4d503a5f9859226a05cfe2a07e46d993
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119313
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 946f457c885bd10ff1a7281c351f3981f035f5a7)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119261
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
|
|
of font attributes, as suggested by Mike Kaganski in
https://gerrit.libreoffice.org/c/core/+/118977.
This partially reverts da59686672fd2bc98f8cb28d5f04dc978b50ac13
but did some modification of the previous code with some
explanationary comments.
Change-Id: I224d9e717bf374a90f4834cbd9e11bf1138b41ff
(cherry picked from commit a91d373837550fbde9c6ba8ca0f43aa8603d86f2)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119118
Tested-by: Jenkins
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
Disables spell-checking on ToC
Change-Id: I6601d8c54974fcc70564292176a71a80d91020e9
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119222
Tested-by: Jenkins
Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org>
(cherry picked from commit 137a1d5380e310a43d36932c643e1331a94fd70d)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119265
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
Regression from commit 862fdb98ca271b60a831cd5420fd16d5f9c1c747
("tdf#143298 Qt5 send SalEvent::KeyModChange events").
Change-Id: I6981af06e3a73dee45ed10416f5d3b17b860a73b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119375
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
(cherry picked from commit 22997c4777eeb424493d1259b72db9a15ff37069)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119267
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
by putting back missing information in lcl_FillRecognizerData
Regression from e98711c14db9348f4d3f7d0f5bbde9276a3e73bc
loplugin:unusedfields in sw (2018-12-12)
Change-Id: I2e99df1a712915851c30c018a1f7279756da72de
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119346
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 717ec99667f5a9ab570f1c8581e2d7a0241c32f6)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119263
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
SfxItemSet::GetItemState returns SfxItemState::DEFAULT without
resetting the passed pointer.
Regression after commit dd5bf12193471f064bf7f581dd1b21783390e735.
Change-Id: I1e427a05f8e5b375f50d81b51d32e78fca83eb42
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119348
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
(cherry picked from commit 0d4cbdbc9cd4ab06056cec66cffd292b41615b6e)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119259
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Revert "use unique_ptr in SdrEditView::ImpConvertOneObj"
This reverts commit 2ec1dab4693cd45b475f02452114dd96ec0bda1b.
I was wrong, we are no returning ownership here.
Change-Id: Ib3b3b32248d4b56f0a48c658db5d6431b26fc352
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119335
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 58c6fad213024ef8f162984a66af565ac9ad4b97)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119260
|
|
Fix regression from commit 759103fd65f82fc131679eba0de05f2ddc7294c8
(tdf#126541 XLSX import: fix View grid lines).
Sheet based grid visibility shouldn't overwrite the
global grid visibility, except for headless mode (i.e.
for command line batch mode, see --convert-to in ./soffice
--help, otherwise we could get a regression here:
The sheet based grid visibility (bShowGrid) is stored
in view settings. Headless mode means not to export view
setting, including sheet based grid visibility.
As the old workaround, use global visibility to keep the
losing sheet visibility.
FIXME: headless mode only works correctly if all sheets
have the same grid visibility. The sheet based bShowGrid
should be moved to another location, which is supported
by the headless mode, too.
Note: now the modified tdf#126541 unit test checks the
fix for the regression of the headless mode, and the new
Python UI test checks the (now improved) fix for tdf#126541
on the desktop.
Co-authored-by: Tibor Nagy (NISZ)
Change-Id: I93551d22cc5d858dc64f93a46cd8aa202b4b2df5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118599
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/+/119314
Tested-by: Jenkins
|
|
Regression from:
https://cgit.freedesktop.org/libreoffice/core/commit/?id=bafd50ee06d982e19d54fae0f9d8f968a2dedbd4
weld DateField Calendar floating window
Change-Id: I43c108908c6e802f3a8014ccf3fa243bebcae0ac
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119293
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
(cherry picked from commit 73503ce975ce0923bfe2cfd075ac8b91de20f29e)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119257
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
The following loop can turn into an infinite loop if the client
disconnects at the wrong point in time, because pSocket->readLine()
returns 0 (indicating disconnection), aLine remains empty (no data
was read), and we just keep looping because we never bothered to check
the return value:
do
{
pSocket->readLine( aLine );
}
while ( aLine.getLength() > 0 );
Aside from spinning unnecessarily, this infinite loop also stops the
server from being able to accept any new incoming connections - in
effect this causes a DOS of the server.
Hence we need to start checking readLine's return value, and break out
of this loop - and in fact break of the surrounding outer loop too
(because we discard this connection and want to wait for another
connection to arrive).
Because of the nested looping it's hard to come up with a clean solution
that only involves changing this loop - we'd probably need to introduce
a temporary to remember that the client has disconnected, and check that
temporary after the do...while - letting us 'continue' the outer loop.
Therefore we first extract the code that handles newly accepted clients
into a separate method, which then lets us simplify the implementation
by returning at those points that we know a client has disappeared. That
unfortunately makes this bug-fix patch a little more noisy than
expected, but it is a refactoring that we'd want to make anyway.
(There are further improvement we can make here, but I'll put those in a
separate commit to simplify cherry-picking of just this fix. Examples
include moving to smart pointers instead of new/delete, introducing an
early return instead of the really long if statement, etc.)
Change-Id: I13c6efa622a1b1de10b72757ea07e5f4660749fb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119083
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit fcc4d8e01360baa9ba0bf20eb5e7a1c9af793a02)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119184
|
|
I guess to get popup windows somehow working, they were matched
to Qt::ToolTip. Fixing that by removing SalFrameStyleFlags::FLOAT
from the tooltip check, resulted in rather bizarre bahviour. The
Qt::Popup window is modal, as documented, but wouldn't close when
clicked outside, like normal popups. And I neither got any mouse
button or focus events, just some leave or enter. The only way
to interact with anything again, was by clicking somewhere on the
popup where it can't take focus... and it has nothing to do with
the transition setting either.
In the end it turned out that Qt::StrongFocus prevented the mouse
events to happen. Even without it the widget accepts mouse and
keyboard focus, so nothing seems lost by omitting that for the
Qt::Popup case; hopefully...
Change-Id: Icda7bbe87268293a400fe0da551d2804675d4c2d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119236
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
(cherry picked from commit 63f92f185b78db5a575da58efac3f94a8cb5a5f6)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119252
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
I originally omitted this in the Qt implementation, as I couldn't
find any non-working use case.
The implementation got a bit hacky, because Qt doesn't have a non-
native way to identify different L/R modifier keys, so we must
process the X11/xkb keycode ourself, which obviously won't work on
Wayland... but most times this is not relevant, so the default
modifier notification may be good enough.
P.S. it's basically the same code then X11 and Gtk...
Change-Id: I37fbfde4a33a966b6017f3e0c280e2c7ea91e4db
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119235
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
(cherry picked from commit 862fdb98ca271b60a831cd5420fd16d5f9c1c747)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119191
|
|
This bug was "unveiled" by commit
963f252cd1ea9c268a6ced68a3454b10cbee1a89 ("Qt5/KF5 get rid
of unneeded own grahics handling"). I'm not sure why I ever came
up with that code. Eventually we would need a paint event, but at
least changing the theme correctly updated LO UI, so that seems
to be handled correctly.
Change-Id: I528b551180be184427eeeeeb2977e2b7da073ce6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119237
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
(cherry picked from commit ec77a2ed0283cb3446f6e352fc329afd3dfb785c)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119190
|
|
Multiple ImplToolItem with mnId 1000 are inserted in the same ToolBox
and the problem is that only the first one gets its mpUserData set,
because setting and getting that is based on the assumption that mnId is
unique per container.
So in Destroy() for every item with mnId 1000 the same AddonsParams
object is deleted.
The problem is that in ToolBarManager::FillToolbar() the local nItemId
is copied, so its value never changes when iterating the loop because
only its copy is incremented.
(regression from ac1aefd5174258d9bcb41465ce566ed1a0819f3d)
Change-Id: Ica47051f9447c2b9d1981f8373d754874f0f31b0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119228
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 7d1175c46588c246c455cd7c8874136bf61197fd)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119186
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
|
|
when the old filter entries are removed in auto-filter rules,
because these options are not visible from the AutoFilter.
E.g. now using AutoFilter doesn't apply copying of the
previous standard filtering, changing also the active sheet etc.
This fixes regression from LibreOffice 3.5 or more.
Change-Id: I26b272dc738f3918ff5b2ca7e8fee3481d21460a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118603
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/+/119272
Tested-by: Jenkins
|
|
on the result area when the "Copy results to:" option is true.
Change-Id: I3da4a1098d8c7c0be6604451b7485312043c5ede
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118484
Tested-by: László Németh <nemeth@numbertext.org>
Reviewed-by: László Németh <nemeth@numbertext.org>
(cherry picked from commit 15684e5d595289e0a94efd80bbe203ca133845f2)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119188
Tested-by: Jenkins
|
|
in a document containing both of them.
Regression from commit 7dd8f8aace536a8e60e87e61ee1d90d61fba15eb
"tdf#120351 DOCX import: fix slow endnote import".
Change-Id: I0fe764f3b48dd2688afa4b7cf0ee6658737ef9fe
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119239
Tested-by: László Németh <nemeth@numbertext.org>
Reviewed-by: László Németh <nemeth@numbertext.org>
(cherry picked from commit 732b08b22eee2682351a9295be29188852fb0489)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119187
Tested-by: Jenkins
|
|
* Fix the wrong calculation of SdrPathObj (line) extent while exporting
to .doc and .docx formats. This resolves tdf#123321 which cause some
lines to disappear after save and reload. This fix does not break
tdf#91687
* added tests to make sure this does not happen again, based on the
.doc file provided in Bugzilla
* Loading the bad exported file in MS Word is OK, but it does not work
in LibreOffice, and some of the lines will be hidden. This needs
additional work.
The regression came from d72e0cadceb0b43928a9b4f18d75c9d5d30afdda
Change-Id: Ic5a7af4d29df741204c50590728542e674ee9f91
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119116
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit bda4d7a7c3fcc259e023f568606be5dcba818db9)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119185
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
We need to acquire the mutex in notifyFinished() to avoid a race between
notifyFinished() and the loop in run(). And we also need to check
mFinishRequested) while holding the mutex to avoid the same race
condition. While we're here, rename the mutex to make it more obvious
that it's not only protecting the queues.
The race can happen because the loop in run() blocks until
mQueuesNotEmpty is set. It also resets mQueuesNotEmpty at the end of
each iteration if the queues are empty. So if someone sets
mQueuesNotEmpty in the middle of an iteration, and the loop later resets
it, the loop won't continue iterating. But we're actually using
mQueuesNotEmpty to indicate either that the queues have data OR that we
might want to finish transmission.
And the problem is that notifyFinished() sets mFinishRequested, followed
by setting mQueuesNotEmpty in an attempt to get the loop to process
mFinishRequested (at which point the loop should finish and return).
But as described above, we can easily get into a situation where the
loop resets mQueuesNotEmpty again (at least if there's no more pending
data in the queues). In this scenario, the loop blocks forever
(we won't receive any new messages after notifyFinished()), causing a
leak of both Transmitter and any resources it's using - e.g. the
StreamSocket.
The easiest way to fix this is to make it clear that the mutex protects
all internal state, followed by using it to protect all internal state.
This issue is not a big deal in normal usage - it's a tiny leak, and
users won't connect enough clients for accumulated leaks to pose
any issues. But it's ugly, and potentially problematic for long-running
tests which could run out of filedescriptors because of the socket leak.
I will rename mQueuesNotEmpty to something more obvious (possibly
mProcessingRequired) in a separate commit.
Change-Id: I2e22087054f3f6a02e05c568b1832ccc5a8b47a3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118751
Reviewed-by: Andrzej Hunt <andrzej@ahunt.org>
Tested-by: Jenkins
(cherry picked from commit 8e6cdb02450a876b5c684eb621e1c9383fb1c428)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118917
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Crash in calc's mouse capture / end tracking.
Change-Id: If5b4cded8ebfc04d1fcf645a7b74184532ab2338
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119034
Tested-by: Jenkins
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
(cherry picked from commit c7b9f477f69c317e0482aeee79b1c2f070a1b981)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119037
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Change-Id: I2dae038d061f036b3b7c0fbbcb6a821305d4839e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119094
Tested-by: Jenkins
Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de>
(cherry picked from commit f5e030a343bec9f0f2c7b30e2dfb21ba5c94e5e8)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119177
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
In addtion, fixed a crash if the start position is greater than the
length of the string being searched.
Change-Id: I9bcda1131324bdfac6957018e91b3a36dd2dc3d6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118996
Tested-by: Jenkins
Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de>
Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119165
|
|
regression from commit I6e97a69c546870199d5a45d9a6ad102e30d820c2
Change-Id: I760dabeb024413f1a2aeb57434f63adedb4246f2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118231
Tested-by: Jenkins
Tested-by: László Németh <nemeth@numbertext.org>
Reviewed-by: László Németh <nemeth@numbertext.org>
(cherry picked from commit d7dff57384126e50b75c95e84fd3081db2f326a1)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119172
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Seen in a crashreport:
handleFatalSignal
./common/SigUtil.cpp:255
/lib/x86_64-linux-gnu/libpthread.so.0
__restore_rt
??:?
program/../program/libswlo.so
SwTextShell::Execute(SfxRequest&)
sw/source/uibase/shells/textsh1.cxx:1496
Change-Id: I709f361a1396abbb07cea06509afc41bb5f5c9a3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119067
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
(cherry picked from commit 801b607fc6b3890eb36d310074f1a68db9ae1f5e)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119173
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Change-Id: Ia2e55886513e81847dd0ac6c697e7cea34e01bdc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119075
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 3ec78de2322f781a98802acc77cfb3e71a08eb3d)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119174
|
|
With CWS impress64 a partial fix for this was implemented to drop
unreferenced named items including all non-color fills after ODF import,
but this is only done in sd so move the code that does that to svx and
call it from sc and sw as well.
Implement some UNO interface for this, it's at least better than a magic
string, and not obvious how a better solution would look like since it's
known only at the end of the import if a bitmap is used or not.
Another problem: when the Area tab is used to change to a different kind
of fill, the items with the details for the previous fill aren't
cleared, and so they are written to ODF files. Hence bitmaps in the
file can be referenced even if they aren't actually used, and bloat up
the files.
Fix this by dropping all unused draw:fill-image-name attributes in ODF
import.
Also do the same for Gradient and Hatch fills; Transparency gradients
can be combined with anything so leave them as they are.
Change-Id: I0b591fd9f963d974d0c3e7208b99621ad61dd93c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118950
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 86c0f58b6f9f392865196606173d1b98a6897f32)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119036
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
|
|
Change-Id: I1997ae2a3916372ded3d4120e4510144c6700777
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119035
Tested-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
with gen rendering part of bt:
2 0x00007fbfb39f440f in __assert_fail_base
(fmt=0x7fbfb3b5d128 "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n", assertion=0x7fbfac2531b0 "!w || dynamic_cast<T*>(w)", file=0x7fbfac2531ca "/home/julien/lo/libreoffice/include/vcl/builder.hxx", line=438, function=<optimized out>) at assert.c:92
3 0x00007fbfb3a03662 in __GI___assert_fail
(assertion=0x7fbfac2531b0 "!w || dynamic_cast<T*>(w)", file=0x7fbfac2531ca "/home/julien/lo/libreoffice/include/vcl/builder.hxx", line=438, function=0x7fbfac3ad73e "T *VclBuilder::get(const rtl::OString &) [T = VclBox]") at assert.c:101
4 0x00007fbfabb27ace in VclBuilder::get<VclBox>(rtl::OString const&) (this=0x6e07e70, sID="DockingFontwork") at include/vcl/builder.hxx:438
5 0x00007fbfabb065e5 in SalInstanceBuilder::weld_box(rtl::OString const&) (this=0x6ce97c0, id="DockingFontwork") at vcl/source/app/salvtables.cxx:6811
6 0x00007fbfb07d3fe9 in SfxDockingWindow::SfxDockingWindow(SfxBindings*, SfxChildWindow*, vcl::Window*, rtl::OString const&, rtl::OUString const&) (this=
0x6e37c90, vtt=0x7fbfafd55730 <VTT for SvxFontWorkDialog+8>, pBindinx=0x34f62b0, pCW=0x6e39c00, pParent=0x34f6f40, rID="DockingFontwork", rUIXMLDescription="svx/ui/dockingfontwork.ui")
at sfx2/source/dialog/dockwin.cxx:774
7 0x00007fbfaf7c3be8 in SvxFontWorkDialog::SvxFontWorkDialog(SfxBindings*, SfxChildWindow*, vcl::Window*) (this=0x6e37c90, pBindinx=0x34f62b0, pCW=0x6e39c00, _pParent=0x34f6f40)
at svx/source/dialog/fontwork.cxx:179
8 0x00007fbfaf7c728c in VclPtrInstance<SvxFontWorkDialog>::VclPtrInstance<SfxBindings*&, SvxFontWorkChildWindow*, vcl::Window*&>(SfxBindings*&, SvxFontWorkChildWindow*&&, vcl::Window*&)
(this=0x7ffc6e31bf70, arg=@0x7ffc6e31bf78: 0x34f62b0, arg=@0x7ffc6e31bf68: 0x6e39c00, arg=@0x7ffc6e31bf80: 0x34f6f40) at include/vcl/vclptr.hxx:279
9 0x00007fbfaf7c1dcc in SvxFontWorkChildWindow::SvxFontWorkChildWindow(vcl::Window*, unsigned short, SfxBindings*, SfxChildWinInfo*)
(this=0x6e39c00, _pParent=0x34f6f40, nId=10256, pBindings=0x34f62b0, pInfo=0x7ffc6e31c230) at svx/source/dialog/fontwork.cxx:169
See full bt here:
https://bugs.documentfoundation.org/attachment.cgi?id=173585
with gtk3 rendering, no crash but:
(soffice:33916): GLib-GObject-WARNING **: 20:40:33.018: invalid cast from 'GtkGrid' to 'GtkBox'
put grid into a toplevel box (git show -w to see minimal diff)
Change-Id: I05f9a295cd839f782467b85d6bafd652a405b14d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118919
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Regression from:
https://cgit.freedesktop.org/libreoffice/core/commit/?id=04aafba860f613c20e7078d038cc83eb02de0b54
loplugin:stringadd simplify some *StringBuffer operations
pulled from a larger patch which I created with a more permissive
variant of this plugin
Specifically here:
-------------- dbaccess/source/ui/querydesign/QueryDesignView.cxx --------------
index 22408bbc58aa..63c5d07998ec 100644
@@ -658,11 +658,10 @@ namespace
if ( field->isAggregateFunction() )
{
OSL_ENSURE(!field->GetFunction().isEmpty(),"Function name must not be empty! ;-(");
- OUStringBuffer aTmpStr2( field->GetFunction());
- aTmpStr2.append("(");
- aTmpStr2.append(aTmpStr.makeStringAndClear());
- aTmpStr2.append(")");
- aTmpStr = aTmpStr2;
+ aTmpStr = field->GetFunction() +
+ "(" +
+ aTmpStr +
+ ")";
}
if (!rFieldAlias.isEmpty() &&
Change-Id: Ib0b57e6c44a58a6a5f9c98aebc6a1213b35108d1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119088
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
Tested-by: Julien Nabet <serval2412@yahoo.fr>
(cherry picked from commit 4a096b9c787fa2079ef1c8f00fed91d432d46fae)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119039
Tested-by: Jenkins
|
|
Change-Id: I1a09193e888318084a7d9b3c76c049e373db545b
(cherry picked from commit 113d308155e4b6a67a8510098a7db5f4a6632bdc)
|
|
don't create when a drawing is selected, wait until its actually dropped
into something (I experimented with dropping into gimp)
Change-Id: I3edaf79537d5723565adf8c45d6f7408026a93e3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118725
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
so select-all and type 'a' will retain cursor at position 1
and
delete/backspace keep cursor position in a semi-rational way
Change-Id: I802f132e58cea9f111deb1c6d157706d08cb5b81
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118921
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
this is a problem since:
commit 67b157e994d9bef01f6117b53fc29e1fee538715
Author: Caolán McNamara <caolanm@redhat.com>
Date: Tue Jun 29 13:00:45 2021 +0100
tdf#143088 multiple Application::EventListeners is expensive
turns out that AddChildEventListener is sufficient to get events
for the widget and its children so the widget-only AddEventListener
isn't needed or wanted
Change-Id: I635419df10fce1d4441b2d85627a6cbe99adb6aa
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118930
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Using plain lang::Locale.Language is always wrong, it may even be
'qlt' for a more complex language tag. As the InfoBar message is
"Please install the hyphenation package for locale “%1”."
actually use the BCP 47 language tag of that character/paragraph
attribution.
Spotted in
https://ask.libreoffice.org/en/question/280102/install-the-hyphenation-package-for-locale-ka/
https://ask.libreoffice.org/en/question/238764/error-message-reads-install-the-hyphenation-package-for-locale-zh/
Change-Id: I5805d4d711989a9d0260940666d3eb33eae830af
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119020
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Jenkins
(cherry picked from commit 64f885a6c9d51999b737c0f61bc3a7fa311a5a94)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118931
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Change-Id: I3e52407a33f79713635063ff4ba374267eb98005
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/119002
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Jenkins
(cherry picked from commit da13b76e07231131cb951868a314ee6f51c0f254)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/118927
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|