Since there already is a one-to-one relationship
between QObject and QAccessibleInterface it makes
little sense to create and destroy the interfaces
on each call to queryAccessibleInterface.
Add a cache and keep created interfaces around for
the lifetime of the corresponding QObject.
This changes the memory management rules: accessible
interfaces must no longer be deleted. If you get an
QAccessibleIntrface pointer that pointer will stay
valid as long as the corresponding QObject is not
deleted.
This also re-enables accessibility for Mac.
We limit the range of the IDs so that they are
useable for Windows directly.
That means we can get rid of the event cache there.
This is based on: Iebf2f374916fc70a9dd29e95f45a6444b85f6cee
Change-Id: I9fe6531812c0dbc5b41101ac05830a6dd75e13a3
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
According to EnumFontFamExProc docs:
> The ENUMLOGFONTEX structure includes the localized name
> of the script (character set).
lfCharSet ENUMLOGFONT's member must be used instead of comparing
these names to non-localized ones.
Also, when the font supports more than a single charset,
EnumFontFamExProc callback is called for it once per charset; thus,
we shouldn't "unsupport" writing systems in a subsequent call.
Task-number: QTBUG-30448
Change-Id: I58fcf32958490cf5a3e873db8335e71a39a9c518
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Check for the existence of a "@2x" file when adding
an image to QIcon. For example, adding "foo.png" will
also add "foo@2x.png" if that file exists.
Change-Id: If32a3446cf56ca0828de17f6a361091e4c874f26
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
We're treating fontDef.pixelSize as a box height everywhere but in
harfbuzzFont initialization code the logic was opposite.
Change-Id: I4eecf7861a6084c7f2dbe1df85a71f2e251da845
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
...by updating HB_Face::isSymbolFont value to make HB know that the font is a symbol font.
U+00AD is a Unicode soft hyphen character, which shouldn't be threated
like a default ignorable (invisible) when 0x00AD occurred in a symbol font.
Change-Id: Idd4760aded644f7f11a1a8d3c03c348abc69a3e4
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Sometimes we get an invalid interface.
In that case the event should just return 0.
This avoids some potential crashes.
Change-Id: I9f166b5d5547dbe7f22c4756502aee811b0b80a6
Reviewed-by: Jan Arve Sæther <jan-arve.saether@digia.com>
QIcon::pixmap() can now return pixmaps that are larger
than the requested size, This is a behavior change
that may require updating existing application code,
and is enabled by setting Qt::AA_UseHighDPIIamges.
Add new pixmap() and actualSize() overloads that take
a window pointer. For 100% correctness you want to
target pixmap generation to a specific window, in
case the system has a mix of normal and high-dpi
displays. (The non-window versions uses QGuiApplication::
devicePixelRatio(), which returns the ratio for the
"best" display on the system.)
actualSize now returns the size in device-independent
pixels - not the pixmap pixel size. This matches
its intended use case which is layout calculations.
Change-Id: I2eefc4ed4aa48b036a7019446694e56213070efa
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
These members are in private classe, so we can safely remove them.
qpathclipper.cpp:762:9: error: private field 'm_point' is not used [-Werror,-Wunused-private-field]
qsimpledrag_p.h💯20: error: private field 'm_cursor_drop_action' is not used
etc.
Change-Id: I35ed486eb959c49bb971700500a66411308ee7bc
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
This was accidentally removed as part of the merge from
dev.
Change-Id: I102ff0d67a136b0dbe467d83edccf02880536d7f
Reviewed-by: Liang Qi <liang.qi@digia.com>
When switching from Xlib to xcb platform plugin it was agreed that
XIM is deprecated. Users should be using QT_IM_MODULE to load input context
plugin for a more advance input method framework support.
The proposed solution is to parse the compose file directly from Qt. This
approach removes the overhead of communication protocols used in Xlib and/or IBUS.
TableGenerator class follows [1].
The compose file is searched for in the following order:
1) If the environment variable $XCOMPOSEFILE is set,
its value is used as the name of the Compose file.
2) If the user’s home directory has a file named
.XCompose, it is used as the Compose file.
3) The system provided compose file is used by mapping the
locale to a compose file from the list in
/usr/share/X11/locale/compose.dir.
Regarding step 3 - TableGenerator searches in hard-coded locations for
system-provided compose files. Here I have introcuded a new environment
variable QTCOMPOSE which can be used to prepend an extra location
to be searched.
[1] http://www.x.org/archive/X11R7.7/doc/man/man5/Compose.5.xhtml
Task-number: QTBUG-28183
Change-Id: I76dcfd454f3acc23db98192a3673c1ab2af4425f
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
1. when there were some engines with ref > 1 in the cache, prior to calling
QFontDatabase::{add,remove}ApplicationFont()/removeAllApplicationFonts()
(QFontCache::clear() has never decreased engine's cache_count);
2. when the QFontEngineData's engine is not in cache i.e. the Box or Test font engine
(~QFontEngineData() didn't free engines it keeps).
Instead of using the font engine's (external) "cache_count" counter,
QFontCache now references a given font engine every time it is inserted to
the cache and dereferences exactly that number of times in clear().
Change-Id: I87677ebd24c1f4a81a53526f2e726e596b043c61
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
As of now, we'll have everything related to the additional formats handling
just in a single place. Make specialData private to prevent accessing it
from outside. This helped in tracking-down several related issues:
- in format(const QScriptItem *), the resolvedFormatIndices can not be empty
at that point, so the code path is dead;
- in resolveAdditionalFormats(), testing if formats has not been indexed yet
is not needed since they are indexed just in the setter;
- in useRawFont mode, hasFormats() didn't check if QTextEngine really
has some formats, which potentially leads to formatting artifacts.
Change-Id: Id4b912888fd5a1fa83f01007170134b6386e2879
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Padding was correctly added between glyphs but it was missing at the
texture edges.
Change-Id: I6d5e1206194f6aecefcfc45ead22d54c1207de4f
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
QOPENGLF_APIENTRY masks over GL vs ES, as we need GL_APIENTRY on iOS,
where APIENTRY is not defined.
Change-Id: I60c097d67e0844c30913c913cf88a9b9e181813b
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
This is mandatory in public headers (qiodevice.h, qopengl*, etc.), but
it's a good idea even in private headers, in case someone includes
that header first somewhere. In particular, all platformsupport API is
private.
Change-Id: If287baa5d9ed14e93c1666efa0e6332c4c1cd9a4
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Being able to calculate advances for series of glyphs include kerning
is important to be able get kerning on QGlyphRun.
Note this kerning is only truetype kerning, since opentype kerning is performed
during shaping.
Change-Id: I8d7458066431cbdce699647056fd0d7a76b20aa2
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
For large fonts, we would get invalid metrics due to the metrics
being too large to fit in the signed chars of the cache struct.
This in turn would lead to the bounding box reported by the
font engine being wrong, and subsequently, the algorithm which
decides whether any part of the glyph is visible in the raster
engine would make the wrong decision, hiding glyphs that should
not be hidden.
Task-number: QTBUG-30026
Change-Id: I59b613ce889fcac0bd50ffec9d369728068d0263
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Ctrl + Shift + A as inverse of Ctrl + A
Works in both KDE and GNOME widgets so put as KB_X11
Change-Id: I3d1781933fcf7db03685453deef8612052cc879a
Reviewed-by: David Faure (KDE) <faure@kde.org>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Kevin Ottens <kevin.ottens@kdab.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Similar to QtWidgets ensure that QGuiApplication::notify() is always
called via its exported symbols. This is needed to allow overloading it
via LD_PRELOAD or other means.
Change-Id: I84d0176b984c5fdf9b199127e08fcfa93dbc853f
Reviewed-by: Harri Porten <porten@froglogic.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Change-Id: I8ecdda35912a95e69c2f8dd98ce9c41c77b222d2
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Remove all trailing whitespace from the following list of files:
*.cpp *.h *.conf *.qdoc *.pro *.pri *.mm *.rc *.pl *.qps *.xpm *.txt *README
excluding 3rdparty, test-data and auto generated code.
Note A): the only non 3rdparty c++-files that still
have trailing whitespace after this change are:
* src/corelib/codecs/cp949codetbl_p.h
* src/corelib/codecs/qjpunicode.cpp
* src/corelib/codecs/qbig5codec.cpp
* src/corelib/xml/qxmlstream_p.h
* src/tools/qdoc/qmlparser/qqmljsgrammar.cpp
* src/tools/uic/ui4.cpp
* tests/auto/other/qtokenautomaton/tokenizers/*
* tests/benchmarks/corelib/tools/qstring/data.cpp
* util/lexgen/tokenizer.cpp
Note B): in about 30 files some overlapping 'leading tab' and
'TAB character in non-leading whitespace' issues have been fixed
to make the sanity bot happy. Plus some general ws-fixes here
and there as asked for during review.
Change-Id: Ia713113c34d82442d6ce4d93d8b1cf545075d11d
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Qt convenience classes for the GL_KHR_debug extension
This allows the developer to
1) ask the GL for a log of the last generated messages;
2) emit Qt signals whenever the GL wants to tell us something.
Change-Id: I830343a26714c51abb68ce1269163c145d1e2aac
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
There is no definition for Yi, Tagalog...Braille in
QFontDatabase::WritingSystem, see the header file.
Change-Id: I0187b0ea23cdb49c4c3306217296d727d4983652
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Desktop OpenGL builds on windows correctly define APIENTRY whereas at
present the ES2 (ANGLE) builds do not. This leads to QOPENGLF_APIENTRY
having the wrong calling convention.
This fix is required for https://codereview.qt-project.org/#change,48660
but may also fix any random crashes that people may be seeing with
other Qt-wrapped OpenGL functions in ANGLE builds.
Change-Id: I8068c181d41be949d29168bd5ca1a181cc2245c7
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
If a line segment isn't filled we keep using the same point of origin.
Task-number: QTBUG-26156
Change-Id: I20af8410a7039b69848f201ab62fd3c01b95531b
Reviewed-by: Titta Heikkala <titta.heikkala@digia.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
The capitalization settings is attempted read via specialData->addFormats,
but the QTextCharFormat there is reset when added, and should instead be
accessed through specialData->addFormatIndices.
Change-Id: I190c419f07c3f7e803ca1d44059e8f538216b9ab
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
This is still the same trivial implementation with the only difference in that
that it properly handles surrogate pairs and combining marks.
This temporarily makes QTextEngine::itemize() insignificatly slower due to
using intermediate buffer, until refactoring is done.
Change-Id: I7987d6306b0b5cdb21b837968e292dd70abfe223
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
This should have been done right from the start,
instead we only made the documentation internal.
Also remove the classes from the BIC data.
Change-Id: I238a7a7cc5d26980b23931c78e7e5a4477d46920
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@digia.com>
Expanding on the change fixing QTBUG-24762 with the realization that any
line needs to be drawn in a consistent way regardless of system or
painter clip, not just dashed lines.
Task-number: QTBUG-25036
Change-Id: Ief7ef19cc92c52e7d792500a581a072ba032767e
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Currently, QCoreApplication::startingUp() returns true even after
a QCoreApplication has been constructed. This patch makes it return
false after it has been constructed and adds checks to QApplication
and QGuiApplication to ensure that it returns true within the
constructor of these classes.
Task-number: QTBUG-2591
Change-Id: Ie511522d35b5658c20be43dd112eae18c205277f
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
For devices taller than 32768, fillRect with linear gradient would
go below the cliprect.
Task-number: QTBUG-26766
Change-Id: I516ede1545909e9ce2de5812c61bb473f99f5759
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Needed as a consequence of I4f5e37480eef5d105e45ffd60362a57f13ec55f5
Otherwise a Window in QML doesn't get notified to update its layout
after being mapped the first time, on Linux.
Change-Id: I8b1291e16b11eb93ab045ac5c5ed6b478391cb21
Reviewed-by: Jan Arve Sæther <jan-arve.saether@digia.com>
This fixes the build after ec9c0faefd (Update gl2ext.h with the
latest version from Khronos).
Just like Mac OS X, iOS is a controlled platform, and we should include
the system headers instead. The OES_EGL_image_external extension is
not yet in the system headers though, so we provide our own typedef
to make the qopenglextensions.cpp code build, and the extension will
work out of the box once Apple includes driver support for it.
Change-Id: Ib6ba09d400ba38f05d91c90d4c9e54a8626889e1
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
HB_FontRec was defined in qharfbuzz_copy_p.h as a void* typedef,
which is incorrect, and is the wrong size. This caused things
accessing QFontEngine data after an HB_FontRec to read from the
wrong offset. The visible consequence of this was incorrect text
rendering in OpenGL in Linux.
Task-number: QTBUG-29966
Change-Id: I71e6a6c5b4ef7b2c8a92631d2b9540537afc56b1
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Avoid using script_to_hbscript() in generateScriptItems()'s loop.
This is absolutely unnecessary due to hbscript_to_script(script_to_hbscript(..))
in QTextEngine::itemize().
Change-Id: I44e710d010f8e57043a6c74dd80f3164e95e18e6
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
This affects HB_Font and HB_Face.
As of now, the Hurfbuzz API usage is concentrated in qfontengine(|_ft).cpp
and qtextengine.cpp, thus it is a lot easier to switch to Hurfbuzz-NG.
Change-Id: Ie06959efd5d6080fe44c407d9f5de0a07dd1c210
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Josh Faust <jfaust@suitabletech.com>
Change-Id: Iea4905d802213848594d2ad0266696e5edb884f8
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
