Our implementation of compose table parser was added on Mar, 2013.
libxkbcommon added APIs for the same thing in Oct, 2014 (ver: 0.5.0).
After removing RHEL 6.6 from the list of supported platforms we were
able to move the minimal required libxkbcommon version to 0.5.0. Now
we can use the xkbcommon-compose APIs on all supported platforms.
With this patch we can drop nearly 1000 lines of maintenance burden.
This patch fixes user reported issues with our implementation.
Known issues:
- Testing revealed that xkbcommon-compose does not support non-utf8 locales,
and that is by design - https://github.com/xkbcommon/libxkbcommon/issues/76
Our implementation did work for those locales too, but it is unclear
if anyone actually uses non-utf8 locales. It is a corner case (work-arounds
existing) and likely a configuration error on the users' system.
- Looking at the release notes for versions above 0.6.1, only one issue
that stands out. Compose input does not work on system with tr_TR.UTF-8
locale, fixed in 0.7.1. Compose input works fine when using e.g. en_US.UTF-8
locale with Turkish keyboard layout.
Note:
With Qt 5.13 we have removed Ubuntu 16.04 and openSUSE 42.3 from CI:
Ubuntu 16.04 - 0.5.0
openSUSE 42.3 - 0.6.1
CI for Qt 5.13 has:
Ubuntu 18.04 - 0.8.0
RHEL-7.4 - 0.7.1
openSUSE 15.0 - 0.8.1
Currently the minimal required libxkbcommon version in src/gui/configure.json
is set to 0.5.0, but we could bump it to 0.7.1 to avoid known issues from above,
but that is a decision for a separate patch.
[ChangeLog][plugins][platforminputcontexts] Now using libxkbcommon-compose
APIs for compose key input, instead of Qt's own implementation.
Fixes: QTBUG-42181
Fixes: QTBUG-53663
Fixes: QTBUG-48657
Change-Id: I79aafe2bc601293844066e7e5f5eddd3719c6bba
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
xcb/eglfs/wayland - all use XKB keyboard configs and APIs. There
is a lot of duplicated and naturally a diverging code. This patch
adds a helper library to avoid all the mentioned problems and unify
feature set between these platforms.
qlibinputkeyboard:
Added a fixup for 2803cdf758. From
spec: "keysyms, when bound to modifiers, affect the rules [..]",
meaning we can't look at keys in isolation, but have to check if
bounding exists in the keymap. This is done by using
xkb_state_mod_name_is_active() API, but that API has its limitations -
https://github.com/xkbcommon/libxkbcommon/issues/88
I will fix this separately in the LTS (5.12) branch.
We need to read the modifier state before the key action. This
patch fixes a regression introduced by aforementioned patch, which
caused modifiers being reported wrongly in QKeyEvent::modifiers().
qtwayland:
Moved toKeysym(QKeyEvent) from qtwayland repository into this library.
For this and other key mapping functionality wayland was duplicating
the key table. All of that will be removed from qtwayland, and calls
will be replaced to use this lib. Adjusted toKeysym() to fix QTBUG-71301.
Qt keys don't map to ASCII codes, so first we need search in our key
table, instead of mapping from unicode.
lookupStringNoKeysymTransformations():
fixed off-by-one error, where we were including terminating NUL in
QString.
Fixes: QTBUG-71301
Task-number: QTBUG-65503
Change-Id: Idfddea5b34ad620235dc08c0b9e5a0669111821a
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
When such modules aren't following the libQt5Foo.so naming convention,
the generated CMake files would be incorrect.
Change-Id: I57908f7466bff7a05f19271ccd495849476bdf38
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Kevin Funk <kevin.funk@kdab.com>
MSVC 2015 has a size limit on strings; sizeof (including the
terminating '\0') must not exceed 0xffff. The generator for the
suffix-list data worked round this by breaking its data into chunks of
at most 0xffff bytes; however, it was limiting on the strlen, not the
sizeof, so was off by one. It checked for this before adding each
suffix, so has (until now) always happened to break early enough; but
the latest update gave an exactly 0xffff chunk, whose terminating '\0'
took it over MSVC's limit. So adjust the cutoff to effectively
include the terminating '\0'.
Task-number: QTBUG-72623
Change-Id: I76ea40060d9fc13c0f7002c5ba22e71b8d0af787
Reviewed-by: Peter Hartmann <peter-qt@hartmann.tk>
Calling a qmake replace function without assignment is undefined
behavior.
This amends 11ae0e77.
Change-Id: Ie716f295275d1ba79a217745b332a8eca04b355d
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
In 71a63836ed a workaround for Samsung
devices using ImhNoPredictiveText was included as there is a bug on
those devices. However this causes a problem with other keyboards such
as Gboard when using languages such as Japanese, as it would not show
the right keyboard at all and only showed a qwerty one in those cases.
Therefore we default to not working around the issue as it is more of a
problem to not allow certain keyboard layouts as opposed to having
predictive text even if it is turned off. For those who want to disable
predictive text and as such have the consequences of not showing some
keyboard layouts, they can set the
QT_ANDROID_ENABLE_WORKAROUND_TO_DISABLE_PREDICTIVE_TEXT environment
variable.
[ChangeLog][Platform Specific Changes][Android] Text fields with
ImhNoPredictiveText set are no longer working around keyboards that
disregard this setting. To enforce the workaround then the environment
variable - QT_ANDROID_ENABLE_WORKAROUND_TO_DISABLE_PREDICTIVE_TEXT
should be set.
Change-Id: I9ace7ba96ebad68987b53783e25067b66c002f25
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
When the application is closed via the task manager on Android
then we need to ensure that that the application can close down
any running event loops. So we wake up all the event loops and
then call quit() directly on the application object to start a
graceful termination of the application.
In order to aid the graceful termination of Qt then a check is
added to ensure that it does not try to create a new surface when
the application is suspended. This prevents it from locking
while trying to create a new surface when this is not possible.
Fixes: QTBUG-70772
Change-Id: I6795b3d280e178d7f1207004a1b965a31a0cc9e9
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
QMake's strategy is generally "pretend everything is Latin 1", which
basically equals "do 8-bit pass-through". Change the handling of
QMAKE_SUBSTITUTES input accordingly to avoid conversion losses when
converting from and to UTF-8.
Fixes: QTBUG-72130
Change-Id: Id903bbd2afa99708c92fd09fab3db944aa819a94
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
QListView::setSelection() algorithm is designed for items to
occupy their cells completely, which is not the case when
itemAlignment is used. The middle part of the selection rect
goes beyond the column borders and extra items are selected.
Use the introduced cellRectForIndex() instead of rectForIndex()
to calculate the middle part correctly.
Fixes: QTBUG-73684
Change-Id: I4a1e42a056d56e85a16d8ae0ffe18b78d1d6deb7
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
If you run an application under eglfs, it falls with segfault on the
exit. For example, examples/gui/analogclock,
examples/widgets/widgets/analogclock, examples/opengl/cube,
examples/opengl/qopenglwidget, etc. (I have added the function
keyPressEvent to exit by qApp->quit(), if needed).
It isn't appear in applications using QQuickView or QGLWindow.
This is because QCoreApplication destructor, where the variable self = 0
(therefore, qGuiApp = 0), is called before than
QOpenGLVertexArrayObject::destroy(), where qGuiApp is accessed
(qGuiApp->thread()).
Task-number: QTBUG-73824
Change-Id: I1dc55d5e811bfe8a8ea2178752e8771f8644d356
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
The standard astc encoder has its own file format.
Change-Id: I9a2f7b1fa20ba344b79637bafb50ff2bd0596747
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
[ChangeLog][Third-Party Code] libjpeg-turbo was updated to version 2.0.2
Task-number: QTBUG-73883
Change-Id: Ib4ed145f1ed9c51933aae565dcb4064ffd60ec24
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
This adds the first uses of Q_DECL_ENUMERATOR_DEPRECATED.
Prompted by API change review for 5.13, comment by Christian Ehrlicher.
Task-number: QTBUG-73484
Change-Id: Ic106048cec18960a91f88eb12650abab6abf0e97
Reviewed-by: Rolf Eike Beer <eb@emlix.com>
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
Regular update in preparation for 5.13, adding tests for additions
since 5.9.4/5.10.1/5.11.0's update 7e946030 (the last to record its
upstream version sha1). Corrected the license header: it's now
published under MPL 2.0 (not 1.1); and our secondary licensing of it
is as LGPL3. Deferred full header over-haul until we've worked one
out in detail.
[ChangeLog][Third-Party Code] Updated DNS public suffix list
Task-number: QTBUG-72623
Change-Id: Iabdbbbfd79624830396c2a6fe0a73389bd6ce5b7
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
The sysrootification of OpenGL include paths must be done only once: at
configure time. The resolved paths are stored since 521a8539 and must not be
resolved again.
Turn the makeSpec-type opengl library into a custom-type one, and do
the sysrootification in the handler function.
Fixes: QTBUG-73736
Change-Id: I2933144057d6f01d8bfc7bda2c2df56c57303459
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Handle WM_SETCURSOR to apply override cursors to window that do not have
mouse capture (as is done in Qt 4).
Fixes: QTBUG-58590
Change-Id: I7ff6f799da1b8d4b4396c0a6137778a11a192617
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Andre de la Rocha <andre.rocha@qt.io>
QSyntaxHighlighter has the follwoing constructor taking a QObject
QSyntaxHighlighter::QSyntaxHighlighter(QObject *parent)
: QObject(*new QSyntaxHighlighterPrivate, parent)
{
if (parent->inherits("QTextEdit")) {
// ...
}
}
Typically, a 'parent' refers to the parent/child hierarchy in Qt
and is allowed to be a nullptr. However, in this case, passing a
nullptr will lead to a crash, as reported in
https://bugs.kde.org/show_bug.cgi?id=404820
This patch makes the QSyntaxHighlighter constructor nullptr safe
by checking if parent is a valid pointer.
Change-Id: Ia4e9b46b94fd37e6ceb2cd0538594353fdc1e349
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Christoph Cullmann <cullmann@kde.org>
If you enter a nested event loop, cause a deleteLater(), exit that event
loop, then enter a new one, the nesting count will be the same so those
are legitimate targets for deletion.
Task-number: QTBUG-73432
Change-Id: Id98140e1c2f0426cabbefffd157f975b5e291ccd
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is required for non-ANSI paths on Windows.
Change-Id: Id98140e1c2f0426cabbefffd157c4065c3bdfd40
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Not entirely sure that this solves the problem reported in the bug
report, but here's the theory: the loop
633 while (!d->interrupt) {
...
710 }
has few calls that recurse back, so the compiler optimizer can assume
that the variable remains unchanged (not interrupted) in most of the
branches. Additionally, it can assume the variable did not change from
there to
712 // still nothing - wait for message or signalled objects
713 canWait = (!retVal
714 && !d->interrupt
715 && (flags & QEventLoop::WaitForMoreEvents));
Which causes canWait to be true, despite having been interrupted by
another thread.
Changing to an atomic does not force the reloading of the variable
(strictly speaking, would need volatile, but all atomic implementations
do reload now), but it solves the problem of data race, which was UB.
The equivalent variable in the Unix event dispatcher is atomic (commit
49d7e71f77 from 2013). I've reordered the
bool members so they're all together and reduce the amount of padding in
this class.
Fixes: QTBUG-72438
Change-Id: I4ac1156702324f0fb814fffd156f290df94dc4c7
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The new test tst_QTouchEvent::touchOnMultipleTouchscreens()
needs the touchpoint IDs to be predictable, but another test currently
has a QEXPECT_FAIL; without release events, g_pointIdMap continued to
hold the touchpoints that were there when the test failed. So it's
necessary to add QWindowSystemInterfacePrivate::clearPointIdMap()
to be able to call it in the test cleanup function.
Fixes: QTBUG-73830
Change-Id: Ia6a70d028be95cd2b6676db6363ec408c0b116bc
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
With a QTreeView it is possible that collapsing an item can cause the
item under the mouse to be a new one and over the checkbox area for the
new item. As a result, a release can cause it to change the check state
even though it did not get the press for that item. This ensures that
it only allows the edit if it got the press as well.
Fixes: QTBUG-61476
Change-Id: I9a0821466afc84c97c9819755ccbacd729f7fbd7
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
It can be used by any platform that uses xkbcommon for keyboard
input: wayland, eglfs.
Change-Id: Iba7d8989f922063cdceb94c1c4bcd8d8e4ae294e
Reviewed-by: Giulio Camuffo <giulio.camuffo@kdab.com>
For creating the qmlplugindump rule we need the name of the import. So
far we're trying to derive this from the directory name, but that
becomes complicated if versions are included, like
QtQuick/Controls.2/Fusion
Instead of trying to tweak the regexp even more to capture this, we
now allow a plugin to set it's name explicitly via IMPORT_NAME.
Change-Id: Ie75dae7d41398b3ac19ccb6910002b6fad009891
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Visual Studio doesn't support files being in multiple filters and
refuses to load such projects. Source files that appear in variables
that are mapped to file filters (SOURCES, TRANSLATIONS, ...) must not
be added to a second filter if they are input for an extra compiler.
Fixes: QTBUG-74004
Change-Id: Id2d752059c98d04e8154a7848c91f29a94bd092a
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
When an item was dragged outside the window, the mouse release event was not
being properly delivered and the item would stick to mouse cursor after
moving it back into the window.
Fixes: QTBUG-72994
Change-Id: Ibce990390c866e16d58f7d969673dd05e862d97e
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Changed DownloadLocation to match the version claimed in Version,
10.32, rather than the 10.31 tar-ball's URL.
Task-number: QTBUG-72623
Change-Id: Id3e17085ea9bb0a854c95e4dce7f472cc38923ed
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qtPrepareTool could return a more complex command-line in the variable,
for some reason. So declare the dependency on the actual executable
only.
Change-Id: Id061f35c088044b69a15fffd1583504f25936a7a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
For extreme coordinates, the rasterizer's width parameter could become
NaN, which compares false with everything and hence would trigger an
assert.
Fixes: QTBUG-56434
Change-Id: I27abae6ab0bc94ce042be86ea0587095cdb7d487
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Fixes the error: variable 'isDifferent' set but not used
Change-Id: Ibd60b17126057da64a41d325b7ef548316f27c4b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This implements conversion parity with QString, which can convert to
and from long and unsigned long.
The implementation simply forwards to existing long long overloads or
uses the existing helpers, so just as for the conversion to/from int
or short, no additional test cases were added.
Change-Id: I37ef06d9ce4d80d98bd72720353996bac723e09c
Fixes: QTBUG-782
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
(cherry picked from commit 783953f09d)
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Say where the git repo is, for finding commit IDs, but make clear that
the download should come from the public site. Revised Description to
the closest match to it I could find on the current web-site (on the
home page). Mention where the script to convert the data live.
Task-number: QTBUG-72623
Change-Id: Ie5f7b2b0c21cdf1c24e311c13866cb1bb02e6973
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Commit 960af0d64 ported away from EM_ASM but also
changed the logic. Set the canvas.style.cursor property,
like it originally did.
Change-Id: Ie4b23abeae54173894bff1db000a305c459684bb
Reviewed-by: Lorn Potter <lorn.potter@gmail.com>
This plugin does not use any XCB libs.
This patch amends 94a4f06fb8
and 744be250bf
Change-Id: I33e2647ace3d9f32a420551b3b198e33a182a06f
Reviewed-by: Michal Klocek <michal.klocek@qt.io>
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Instead of hardcoding the target's extension to ".exe" we should rely on
target information available in Visual Studio. $(TargetFileName) is
documented as "The file name of the primary output file for the build
(defined as base name + file extension)." so it can be used instead of
$(TargetName) together with ".exe".
Change-Id: I103d8d13456910617b2d53c9c8f4e2935eb93015
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
We must deleteLater the pipe writer in closeChannel, because if you
call closeWriteChannel() from a slot that is connected to a signal
emitted from QWindowsPipeWriter, we'd operate on a deleted object.
For consistency, we're calling QWindowsPipeWriter::stop before
deleteLater and deduplicate the code.
Fixes: QTBUG-73778
Change-Id: I61d3dedf57e9fd02517a108d13ffc85e006330f6
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
Wrong locale name
-----------------
From the java.util.Locale documentation:
"Note that Java uses several deprecated two-letter codes. The Hebrew ("he")
language code is rewritten as "iw", Indonesian ("id") as "in", and Yiddish
("yi") as "ji". This rewriting happens even if you construct your own Locale
object, not just for instances returned by the various lookup methods.
Because of this, if you add your localized resources in for example values-he
they will not be used, since the system will look for values-iw instead.
To work around this, place your resources in a values folder using the
deprecated language code instead.
See also:
http://www.apps4android.org/?p=3695https://issuetracker.google.com/issues/36908826
Change-Id: I726c43f282156b21e8d6b073029f3c3b8fd42a30
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
