| Age | Commit message (Collapse) | Author |
|---|
|
The bugdoc is similar to cool#11785, i.e. a document of 3-400 pages, but
the content is reasonably simple, mostly plain text and bullet points.
The document has a shape, though. Loading this document in COOL does a
full sync layout on load, which is reported to take ~9s on a slower
machine and it takes >1s for me locally (after doc load, before the
render of the first page).
It seems what happens is that idle layout is started, but it's not
interrupted, because moving content between pages in
SwLayAction::Action(), called from the SwLayIdle ctor calls into
SdrPaintView::Notify(), which starts a high priority "idle", so the LOK
client thinks it should not interrupt core jobs, so the idle layout
first completes, and only then we paint the first tile.
Fix the problem by adjusting the priority of
SdrPaintView::maComeBackIdle dynamically. Add a flag in the draw model
to know when we're inside the Writer idle layout and once the svx/ idle
is started, lower the priority in case that's invoked from inside the
Writer idle layout. This gives the LOK client a way to interrupt the
Writer idle layout (via the anyInput callback), now that it seems core
has no high priority jobs in the scheduler. Size of the time window
between the end of doc load and the start of the first tile render,
before: 1617 ms. After: 182 ms
An alternative approach I considered is to interact with the
SdrPaintView that belongs to the current Writer view, but that won't
work. There are typically 2 draw views active right after document
load, one created in SwViewShellImp::MakeDrawView(), but there is also
an additional hidden view created in the SvxDrawPage ctor for UNO
purposes, so if we don't go via SdrModel, we would have to notify both
views that we're entering / leaving the Writer idle layout, which looks
more complicated than what this patch does.
Change-Id: I115ba2c152fe7283804cde564afd511f9fbff707
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185360
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
Implements a minimal functional version of ruby character layout and
rendering in Edit Engine. Many features are missing from this
implementation, and will be added in future commits.
Currently, ruby characters cannot be created via the user interface.
Ruby character styles are also currently hard-coded.
Change-Id: Ia7c18ccae6c37cd320acc214574ba3123c18ddb0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185210
Reviewed-by: Jonathan Clark <jonathan@libreoffice.org>
Tested-by: Jenkins
|
|
JSDialog API cannot communicate without vcl id of a widget
Change-Id: Ic6bd0d430207f32cf631564102511eae7130b7bb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184721
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185348
Reviewed-by: Szymon Kłos <szymon.klos@collabora.com>
Tested-by: Jenkins
|
|
Add a config header for this, and split yrs.hxx in two because the
actual libyrs.h is only needed in 2 files.
Change-Id: I30453fda3082f977e328ed56a4a101743ecdd586
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185329
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
XWindow2 subclasses XWindow, so there's no need for
a VCLUnoHelper::GetWindow variant taking an XWindow2
reference that does the same as the variant taking
an XWindow reference.
Change-Id: I6f931252d2ae2e046dc31c076094e71b8aa71346
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185273
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
TL;DR
Now there is only one entity "THEME" the user has to care about when
using the appearance tab page. We ship three default themes which
just change the document colors. Custom themes can customize the UI
too.
We used to have two sections in the appearance tab, the themes dropdown
and the appearance radio buttons group (automatic, light and dark), and
this was really confusing.
Let's say I am a new user, and I go to the appearance tag page I select
the dark radio button, why isn't my application changing to dark mode?
same for light... and then the distinction between application colors
and document colors, the radio buttons were mainly for the docuemnt
appearance, and the themes dropdown was for changing the current theme
which might change the document colors as well.
Now the radio buttons have been merged with the themes dropdown. Now
there are three default themes there, "Automatic", "Light" and "Dark".
These three themes don't define any UI colors and just change the
document colors. If the user installs a new theme, and if that theme
defines UI colors, the UI will also be customized.
Change-Id: If8da6d2f729f117ada07e13b607ed8525395cb6c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184429
Reviewed-by: Sahil Gautam <sahil.gautam.extern@allotropia.de>
Tested-by: Jenkins
|
|
Based on the Writer's HTML export code. For now, it doesn't export
anything yet, just creates an empty file.
Change-Id: I0d484d2b5bc549f34e674c8f858acb352e275c83
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184201
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 8e83a343a49b2053bc08326a109956f224a47ae1)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185283
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
The weld::Widget API for grabbing is inspired by
`gtk_grab_add` from GTK 3 [1] which grabs both,
mouse and keyboard, blocking interaction with
other widgets in the application.
For LibreOffice, only grabbing the mouse is of
interest. CustomWidgetController::CaptureMouse
is the only caller of weld::Widget::add_grab,
and the VCL implementation in SalInstanceWidget::add_grab
also only calls `Window::CaptureMouse` on its
vcl::Window.
In addition, there is a concept of a "grab count",
i.e. calling `weld::Widget::grab_add` multiple
times without `weld::Widget::grab_remove` means
that `weld::Widget::grab_remove` needs to be
called the same amount of times later on.
However, the concept of having "multiple grabs" at
the same time is also not used anywhere in LO
in practice, and the VCL implementation also
doesn't implement that concept.
Therefore:
1) Drop the logical concept of multiple grabs from the
weld::API. Adjust the gtk3 implementation
to make sure to have at most one grab at a time
that is owned by LO application code.
(Internal GTK toolkit code may use additional grabs.)
2) Rename the methods to make clear that only
the mouse grab is of relevance.
(The gtk3 implementation still uses
`gtk_grab_add`, i.e. grabs the keyboard
as well.)
In GTK 4, there is no more API for grabbing the
mouse and only a counter was used to keep track
of how often weld::Widget::add_grab had been called.
Switch that to a bool instead.
Qt provides separate methods for grabbing the mouse
(QWidget::grabMouse [2]) and the keyboard
(QWidget::grabKeyboard [3]).
This commit is also to clarify that only the former will
be needed when implementing this weld API for the Qt-based
VCL plugins in an upcoming commit and grabbing the keyboard
isn't needed.
[1] https://docs.gtk.org/gtk3/method.Widget.grab_add.html
[2] https://doc.qt.io/qt-6/qwidget.html#grabMouse
[3] https://doc.qt.io/qt-6/qwidget.html#grabKeyboard
Change-Id: If6f34a5eed8efe3dfb8d453c1ee97796cd4418c7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185254
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
Change-Id: I6f661515e2de11ea1ead0b79435413d6a0e81c9d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184264
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185142
Tested-by: Jenkins
|
|
This helped to write the tdf#166229 testcase, because it shows where the
actual bullet character is.
Also declare the encoding in SwDoc::dumpAsXml() explicitly, since these
bullet characters are typically not ASCII, so they didn't show up
correctly in the output.
Change-Id: I5a6f58d0971c2542e4cd79209589259fa5e9e5b7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185243
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
Change-Id: Ib6f3b732a7eb6f2dbf423ffd2206b0474ff6c681
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185200
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Replace unreliable integer cursors with WeakRef cursors which maintain
their position automatically.
This requires yrs v0.23.1 with the yweak_read() function.
Unfortunately this crashes currently, so hide it behind #if YRS_WEAK.
https://github.com/y-crdt/y-crdt/issues/536
Change-Id: I74ec1eaa25c7cb554efd38ee632df041ac92d5c2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/183311
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
The conversion serves no goal, and no place benefits from that.
Change-Id: Id2f1865fb8ef0cdb51f959b35e482f35fad8b75f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185240
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Tested-by: Jenkins
|
|
Since commit ef8341f05ec79fcdac3b05b45659f5c98ff05edf,
XFilePicker3 is used there; XFilePicker2::getSelectedFiles
is guaranteed to be available. Fallback to deprecated API
is not needed, and this allows to drop some ugly code.
Change-Id: Id11a5238f3db5930c30f80df6ee6a77a377f6475
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185232
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
...warning about non-overriding virtual dtors in final classes. Every single
one of these might or might not be suspicious (e.g., in some cases it might be
better to have a virtual dtor in a derived-from class). Also, for some of the
virtual dtors, making them non-virtual then caused
> error: unnecessary user-declared destructor [loplugin:unnecessaryoverride]
so those were removed completely.
Change-Id: If0ee71645a347d6096b65988c56f627c42fe54b2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185223
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
|
|
Being a weld::Toggleable subclass, weld::RadioButton
implementations need to implement
weldd::Toggleable::{g,s}et_inconsistent.
However, neither the VCL implementation (SalInstanceRadioButton)
nor the Qt implementation (QtInstanceRadioButton)
have a concept of an inconsistent state for radio buttons
(while checkboxes/check buttons do).
The SalInstanceRadioButton implementation was
already doing nothing when trying to set the
state, and always returning false when
querying the state.
Move the logic to the base class, weld::RadioButton
and add an assert in weld::RadioButton::set_inconsistent
to prevent any users from trying to set that state (to true).
(If there any existing ones, they need to be adjusted.)
Setting the state to `false` (as e.g. implicitly
happens when calling weld::Toggleable::set_state with a value
of TRISTATE_TRUE or TRISTATE_FALSE) is fine, only trying to
set to `true` is problematic.
For GtkInstanceRadioButton, explicitly call the new
weld::RadioButton versions. (Previously, the ones
derived from its GtkInstanceCheckButton base class
were used.)
Change-Id: I292ff838b62205d906fb2faf0d391c503a4786e0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185214
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
Previously, it stored the namespace alone; and added ":" each time
the value was used, when it was non-empty (either in GetNamespace,
or when passed to svt's HtmlWriter).
This change only adds the colon once, when maNamespace is set.
Change-Id: Ib16586ebbf99bea7d301d3ac7a870ff4e68ee50b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185203
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
Replaced hijacked Okay and Reset buttons by special buttons
Change-Id: If8315be3f8f71ca1fc5fae0136eef4a76232b3e8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185219
Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org>
Tested-by: Jenkins
|
|
Change-Id: Iedd79b069c145de4c3b70d50552df8bf7da66a1e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185218
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
Change-Id: I57a19add67b4ffff12fa8af575d997bcda378a4e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185206
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Jenkins
|
|
The cursors are represented as integers in yrs as a first step, which
isn't ideal.
Surprisingly EditView doesn't really have functions to correct the
cursor, only ImpEditEngine::UpdateSelections() for the case when nodes
are deleted, apparently there's only one active EditView at a time
typically so it wasn't needed? Typically the active EditView's cursor
is assigned from a return value of ImpEditEngine.
Add some new functions to correct peer cursors and call them all over
ImpEditEngine; unfortunately UpdateSelectionsDelete() isn't easy to get
working on EditView's actual EditSelection cursor so leave that for the
future...
To fix EE paint of non-active comment, add SwAnnotationWin::Paint()
To get things to paint with "gen", it's enough to invalidate the
SwAnnotationWin, but for gtk3 a queue_draw() must be called, else it's
only painted when clicking on the window.
Cursors are painted in color in SidebarTextControl::Paint()
Change-Id: Id8696238290405554ece1b9961676972a559b58e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/181679
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
See tdf#94879 for motivation.
Change-Id: I626e35f8dd44ef4fd8311eb2312db92a0ebff41b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185192
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
No need to lookup the element in this - we already have the iterator.
Introduces SfxItemSet::GetItemState_ForIter for the task.
Change-Id: Ib85706201c452b5345df411c317e79bcf6c17238
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185167
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
With tabs for introduction (and later perhaps also as what's new
info), appearance, and UI picker.
Run ./soffice --safe-mode to test but make sure to start the app
next in normal mode
Change-Id: Ia12ba44c8796fb53f2ec1dbaf46dd25a6ec113cb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184195
Tested-by: Jenkins
Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org>
|
|
Previous changes up to
Change-Id: I7d44e8a8907fa3946d9b9a249e12c542fa656f1c
Author: Michael Weghorn <m.weghorn@posteo.de>
Date: Sat May 10 00:47:02 2025 +0200
gen: Port SystemEnvData::{aShellWindow -> pSalFrame} for DnD
replaced all uses of SystemEnvData::aShellWindow,
so drop it (and adjust code that was previously setting it).
Change-Id: I60d69c6e5cb8b7bb9343d419ea6a6ea2a4fc1d22
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185123
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
|
|
This is a subtask of tdf#165742: Chartex charts are lost on input from OOXML
and re-export.
Fix various aspects of input and output for chartex (and specifically
for funnel charts). This now allows for loading and re-exporting a
very basic funnel chart, with the result that MS Office successfully
loads the exported chart, and it looks similar (though not quite
identical) to the original.
Change-Id: I6eeb277e31250031604f7cdd21b80848a9c642ca
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184758
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Tested-by: Jenkins
|
|
Change-Id: I918b46c29791f865e841eb967b2799f7986538bc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185027
Tested-by: Jenkins
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
Tested-by: Xisco Fauli <xiscofauli@libreoffice.org>
|
|
This was supposed to set some kind of "default directory" for file pickers.
It was never implemented for anything except the "gen" vcl plugin - and
is also not really needed.
Change-Id: I94c6ff04984fc7b1d600530dc26bd8a7672646cd
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184064
Reviewed-by: Samuel Mehrbrodt <samuel.mehrbrodt@allotropia.de>
Tested-by: Jenkins
|
|
Added font panel for chart text based elements:
Title, Label , Legends, Axis.
Font sidebar updated based on the selected chart element properties.
Clicking on the font sidebar will change the properties of the
selected chart element. (Bold, italic ...)
1 title can have multiple Font settings, the sidebar will display
a values only if the whole title havbe the same value.
Setting a value will set it for the whole title at once.
If a Title is edited, then the font panel is disabled now, because
it need an other implementation, to set textEdit attributes,
and to update the font panel based on the attributes.
Added 2 //Todo: where it could (started to) be implemented.
Added 3 new Context to EnumContext:
- ChartLabel
- ChartLegend
- ChartTitle
This is a bit risky because what was previously only a Chart, now
may be not a Chart value, but instead 1 of the 3.
refactored those parts where i seen it could be a problem,
i hope i not left out something.
(Axis was already made by someone)
Change-Id: I19eaeb93332a08faae6ed21e7e81a9ea56a4495b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/185023
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
Change-Id: I76268d39a8792c87a8fcad23dc3f393c4c596486
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/183192
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
This can be enabled in configure via --with-yrs=...; see also README.yrs
Currently this is hardcoded to run over a local named pipe.
Everything related to editengine is implemented in EditDoc and made
accessible via its wrapper classes; everything related to communication
and accessing sw's comment is implemented in sw::DocumentStateManager.
The acceptor starts in SwView::SwView(), once SwPostItMgr exists.
There is a hack in SfxFrameLoader_Impl::load() to fetch the document
that the accepting soffice has loaded, and load it in the connecting
soffice as well.
Change-Id: I89476b5864b70f479bcf15989374c1c65b5da9ea
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/175652
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
Instead of using EditView's Paint directly to Window I
changed this now to using the Overlay that is available in
all applications. This allows to have less repaints of the
Calc View since the Overlay has a copy of the area it is
working overlayed.
It also allows to get away from one of the last areas where
XOR selection was used since EditView/EditEngine have no
better way to do this. Selection is now visualized using the
OS'es choosen SelectionColor, as in all other apps.
I could also get rid of the flush() of the OverlayManager
that was needed before and caused some problems (see
tdf#165621 and associated tasks).
A failing UnitTest showed that OverlayManagerBuffered needs
to be aware when no CompleteRedraw was done yet and thus no
background data exists and refresh makes no sense. This may
happen when there is no redraw/UI e.g. UnitTests, but still
an OverlayManager is used and together with a Reschedule
(many possible reasons) that refresh might be triggered.
Made access to EditView in OverlayObject no longer dependent
on ActiveView, that is not always the one with the EditView.
Now using a combination of ScTabView and ScSplitPos for
accesses. Asserting now if no EditView -> is mandatory.
Continuing going through CppunitTest_sc_tiledrendering UTs,
one more identified.
Have to re-activate EditViewInvalidate rect forwarding
since the Invalidate is needed for sc tiled rendering
tests, but not for sd tiled rendering tests. Added that
to the impl of EditViewCallbacks using
LibreOfficeKit::isActive().
Change-Id: I41f5448bfbabfaae4858c7617478771053c2cd77
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184830
Tested-by: Jenkins
Reviewed-by: Armin Le Grand <Armin.Le.Grand@me.com>
|
|
See tdf#94879 for motivation.
Change-Id: Ie9087a753e871eb684b8dbec2e74a89b7f0f3998
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184950
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
|
|
we do not need to go via this complex luminance calculation, we
just want the red channel of the RGB data.
This is also what the skia backends do, ever since
commit 2fcfbd73768b69ba58607a054e7f851be2942992
Author: Luboš Luňák <l.lunak@collabora.com>
Date: Fri Apr 3 22:50:12 2020 +0200
no gray conversion needed for VCL alpha hacks
Change-Id: Ie4a7adcc7c488d241ec58e64a130da544c3d39d0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184944
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Jenkins
|
|
ApplicationAppearance registry entry holds the appearance mode the
application is in at any point. it is represented by three integers 0
for "Automatic", 1 for "Light" and 2 for "Dark".
this patch introduces an enum `AppearanceMode` which replaces the
integer switch cases and comparisons in conditionals with more readible
enums.
Change-Id: Ibe90875cc8d2370476442f726ebbb4eb02cccc91
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184393
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
|
|
Integrate build system support for the Zstandard (zstd) library and
implement the corresponding C++ decompression stream classes, enabling
future use of zstd-compressed streams in ZIP packages.
Build System (Part 2a):
- Add `--with-system-zstd=[yes|no|auto]` configure option.
- Implement system library checks (`zstd.h`, `libzstd`) in configure.ac.
- Define rules (`external/zstd/*`) for downloading and building the
internal zstd static library (decompression components only).
- Set up necessary CFLAGS/LIBS via configure.ac and config_host.mk.in.
- Configure linking via RepositoryExternal.mk (`gb_LinkTarget__use_zstd`).
- Update auxiliary files (download.lst, distro-configs, etc.).
C++ Implementation (Part 2b):
- Implement `InflateZstd` (in `include/package`) and `InflaterBytesZstd`
(in `package/inc`) inheriting from `Inflater`/`InflaterBytes`.
- Use the zstd streaming API (`ZSTD_decompressStream`) following the
existing single-call pattern established by the zlib implementation.
- Manage context (`ZSTD_DCtx`) and input buffer (`ZSTD_inBuffer`) state.
- Handle errors, stream completion (`ret == 0`), and input truncation.
- Add new classes to `package/Library_package2.mk` build.
This completes the infrastructure for zstd decompression support.
The next step (Part 2c) involves adding logic to select the
appropriate Inflater based on the ZIP entry's compression method ID.
Change-Id: Ia673f3f19b6a751ba5225394fcf8c3f145c4ae63
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184243
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
|
|
Move the a11y classes used for implementing Basic IDE a11y
from vcl to basctl. Move them into a namespace
"accessibility" to avoid a name clash with the Paragraph
class declared in include/editeng/outliner.hxx.
In TextWindowPeer::CreateAccessibleContext, call the
Document ctor directly and drop the now unused
AccessibleFactory::createAccessibleTextWindowContext.
Change-Id: I396507e5fca17b4e7a1b65b28797569bc31d6a04
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184838
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
Tested-by: Jenkins
|
|
...after d0ee08cfbf12145027eee7ad46448a8734693c06 "oox: export embedded font
used in a presentation document" caused CppunitTest_vcl_font_ttf_structure_test
to fail with
> vcl/qa/cppunit/font/TTFStructureTest.cxx:47:5: runtime error: member call on misaligned address 0x7de9211e0262 for type 'o3tl::sal_uInt32_BE', which requires 4 byte alignment
> 0x7de9211e0262: note: pointer points here
> 00 00 00 00 80 00 00 af 10 00 20 48 00 00 00 00 00 00 00 00 57 33 43 00 00 40 00 20 f0 02 03 20
> ^
> #0 in (anonymous namespace)::testReadTTFStructure::TestBody() at vcl/qa/cppunit/font/TTFStructureTest.cxx:47:5
Change-Id: I4d46decd497fccf7eecdb38764ac82ed2a68b5d6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184829
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <stephan.bergmann@allotropia.de>
|
|
The pre-ODF 1.5 encryption necessarily destroys macro signatures, but
the user would learn this only when re-loading the encrypted document.
Change-Id: Ic8a0190205412f484f47a4b8d864b280eea0d23c
(cherry picked from commit 558c6a83c210a4d9d7a699616c6f254fb23d7cf4)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184822
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
|
|
separately from all private info files during save.
The new expert config option name is: RemoveEditingTimeOnSaving
(There wasn't any UI element added for this expert config option.)
Change-Id: Iab69d062d161ef6aee1c3eb87ada1c16f8b4fd41
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184789
Tested-by: Jenkins
Reviewed-by: Balazs Varga <balazs.varga.extern@allotropia.de>
|
|
This is needed to prevent that we embed a font but then export
a different (substituted) font name into the document, which
would defeat the purpose of embedding in the first place.
This is not ideal either as a we probably need better control of
what we can just substitue and what we need to embed.
Change-Id: Icd29789ceb739b6950576379358a96cc0449b3b1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184696
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
This adds support to export the fonts that are used in an
presentation document. The fonts are converted to EOT fonts.
Adds an EOT converter, which converts a TTF font to EOT font by
writing a EOT header and adding the TTF font data.
Adds BigEndianTypes so we can read a structure that is defined in
big endian and the types get converted (if needed) to system
endian when accessed. + test
Adds TTF reader which reads the true type font structure in an
easy way that is used by the EOTConverter. + test
Change-Id: I8b0bcb69e47943c2a8679edf881d7c2bab45c450
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184288
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
The base class implementation in
OAccessibleComponentHelper::getAccessibleIndexInParent
already implements the same logic.
Change-Id: I3b9e4b2aa04bb24bb0355897ea18eb41cad9ef61
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184794
Tested-by: Jenkins
Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
|
|
Change-Id: Ia03404a18a12aeae829d9358700dc8c10ed4b0b2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184782
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Tested-by: Jenkins
|
|
Refactor ZIP decompression classes (Inflater, InflaterBytes) to support multiple decompression algorithms, enabling future Zstandard
support.
Make Inflater and InflaterBytes abstract base classes defining a common interface. Move the original zlib/Deflate logic into new derived classes InflateZlib and InflaterBytesZlib.
Update usage sites to instantiate the *Zlib variants via unique_ptr to the base class, preserving existing functionality. Move header files to standard locations (include/package, package/inc) and update includes and build files.
This pure refactoring prepares the codebase for adding Zstandard (zstd) decompression. Correct a minor logical bug in the ZipFile::getSizeAndCRC loop condition.
Change-Id: I9be3c02946e29eb2dae09f500eb47d8a4261af22
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/183504
Reviewed-by: Devansh Varshney <varshney.devansh614@gmail.com>
Tested-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
Tested-by: Jenkins
|
|
there is no DeInitVCL in fuzzing, so if the BufferedDecompositionFlusher
thread is started nothing causes it to exit before _exit.
Change-Id: I62463ce8126a0cf0c67f4218bdf66a140f3a021d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184731
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolan.mcnamara@collabora.com>
|
|
This is a subtask of tdf#165742: Chartex charts are lost on input from OOXML
and re-export.
Fix some incorrect URLs and ordering of elements in a <xsd:sequence>.
Change-Id: I4092e91851e80114d3cb8e22f602c33faf146969
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/183736
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
Embed fonts depending if the attribute "embedTrueTypeFonts "on
"presentation" element is set to true. If true, set the doc.
setting "EmbedFonts" and start proceed with font embedding code
path.
Change-Id: Iba2552ab58c07cfa41afca4e1883ae706eb8c678
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184403
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
This is a subtask of tdf#165742: Chartex charts are lost on input from OOXML
and re-export.
Build out the some of the export structure for chartex charts. This now
produces an output chartEx*.xml file with no c-namespace content, for
funnel charts only, and what appears to be fully well-formed OOXML
per the office/drawing/2014/chartex schema. However, MSO doesn't load
the result successfully, so I'm calling it step 4.1.
For the bulk of the output code in chartexport.cxx, there was a design
decision whether to (1) have a single 'if (chartex) ... else ...'
branch point high up, or to (2) branch much more granularly within the
code. (1) has the benefit of cleaner reading, without so many ifs, at
the expense of a lot of duplicated code. (2) allows for much less code
overall but with lots of ifs. For the time being I've taken approach
(2), not necessarily because it's the right one, but mostly because it
is much easier to move from (2) to (1) (largely by just duplicating
the code and in each large branch deleting the local branches that
don't apply), than from (1) to (2) (which would require examining the
differences between the two large branches in detail, in order to re-
integrate the code).
Change-Id: I78de1484359c358b99320b417ab68cd90dab3e55
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/183561
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
See tdf#94879 for motivation.
Change-Id: Ic4cd593ee3605e2c6194ec52e25d27e0e8455b93
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184693
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
