tst_QtEndian's transformRegion_template() was getting a
signed/unsigned comparison warning when T was unsigned in a
QCOMPARE(T-value, 0); so use T(0) instead.
Change-Id: I78cb2ab96f79393def65ed2c020aa3039017ab92
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Replace lines of formatted output with dictionaries from which that
output can be generated efficiently (in sorted order). Tidy up the
python code, in the process.
Change-Id: Ifb74eb69342646c5dbabc3247b3a8ddef602e934
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The scripts that generate the C++ enum names clean out spaces, so
there's no need for enumdata.py to use the squished names; instead,
include spaces where they belong and use them in the relevant names
where they show up.
Change-Id: I7261c87d69e654bc2e89f44d7ebbc06bfb403c58
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
That way, we'll get told all the cases that fail, rather than only the
first. Provoked by investigation of failures that turned out to be
caused by QTBUG-69875.
Change-Id: I8fa2902cbbcb307cbe1fdec2e7d8d6b0c3eb998a
Reviewed-by: Erik Verbruggen <erik.verbruggen@qt.io>
It was added 558dd7f587
It was missed after 128a6eec06
Change-Id: I99f5c42ea45227bef08092e22d656e2d7d3cf60f
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
It's unclear what the original code was doing. It relied on 'm_release'
which could never be 'false' (ref. QTBUG-69679). It was subtracting event
times and comparing with arbitrary '10'. On X11 auto-repeat keys can be
detected by checking time and keycode of the current release event and
the next event in the queue. If an event is an auto-repeat, then next event
in the queue will be a key press with matching time and keycode.
Verified that auto-repeat was unreliable in Qt 4 as well. With this patch
auto-repeat works as expected.
Added support for Xlib's XPeekEvent in our XCB implementation QXcbConnection::checkEvent():
"The XPeekEvent() function returns the first event from the event queue,
but it does not remove the event from the queue."
Sneaking in one variable renaming: "string" -> "text", to match the QKeyEvent::text().
Task-number: QTBUG-57335
Task-number: QTBUG-69679
Change-Id: I0a23f138287f57eaaecf1a009bd939e7e0e23269
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Few lines above we use the same attributes to create the window.
If attributes don't change, there is not need to call
xcb_change_window_attributes()
Change-Id: I1335b8be866bdd9911d7911c87f09091bf77e955
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
... to check for buffered events. This makes the code less verbose and
easier to read. Changed the filter signature to pass an event type in addition
to the actual event, for the convenience of API user. And do not pass worthless
nullptr-s to the filter.
The only reason why KeyChecker from qxcbkeyboard.cpp was not converted
to lambda expression is that the code looks suspicious - KeyChecker::m_release
default value is 'true' and I don't see where it would ever be assigned
'false' (ref. QTBUG-69679) and the code is known to be buggy (ref. QTBUG-57335).
Those issues are out-of-scope for this patch.
Change-Id: If2fdd60fbb93eb983f3c9ad616aaf04834fede9f
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Mikhail Svetkin <mikhail.svetkin@qt.io>
Respect allowRootCertOnDemandLoading, as it's done in QSslSocket (well,
almost as in QSslSocket).
Change-Id: Ic6cbb24a91e92cdb20f5f749553f15a62aae8b02
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Previously, the function had a different name that made its purpose
clear - "abort after peer verification error was encoutered". Since
now it's just 'abort handshake', it also should abort an ongoing
handshake, even if no peer verification error found so that we
now have an API that can reset a QDtls object to its initial
'nothing done yet' state.
Change-Id: Idadfec6f82d65c8f07d1c2afa4467c921c7e85c4
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Change-Id: Id0118e44e1ed44fecdd803dbc57ca0588a1afec0
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: Martin Smith <martin.smith@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
We now include the phase, which is helpful, and we distinguish between
modern wheel events with a pixel and/or angle delta, and compatibility
events sent from QWindowSystemInterface::handleWheelEvent with the old
Qt 4 single value delta and orientation.
Change-Id: I2efe9fe7a55c43240ec94bfa70cafaac6d52603c
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Expands the existing conversions for ARGB32<->RGB30 to also handle
RGBA8888 to avoid any performance regression of the 16bpc conversion
path.
Change-Id: I34e519a04276b0e7ff00c6125ba06c889045d2ac
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
This update corrects several minor documentation errors that
were introduced by a recent change to QDtls.
Change-Id: I6ee6b0bf74c82dca86ba8d5f39acbd642829ec44
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
This reverts commit 823acb069d.
It caused some test failures in qtdeclarative and etc.
Task-number: QTBUG-69891
Change-Id: I2e4038a46de254834e6389c63f6dad0c2e523b8e
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
This caused a crash on destruction because as soon as you construct a
QTouchDevice it will register itself to a list of devices. On application
exit the function cleanupDevicesList() in qtouchdevice.cpp would go
through all registered QTouchDevices and destroy them. Therefore, there
is no need to delete the QTouchDevice from QWindowsPointerHandler.
This was a regression that was caused by
20d6dac63c
Change-Id: I58fb50016c047c3843a3f9677f2c2ef824223d43
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Andre de la Rocha <andre.rocha@qt.io>
While the property is deprecated, we should still set it, even in the
legacy free constructor. The window system never sends x and y changes
at the same time, so either coordinate will be null, therefor add them.
This should restore the old behavior.
Task-number: QTBUG-69089
Change-Id: Ie329cb4196f53c70b25934cbbac5e2e48c65d2b3
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
It is not relevant anymore. C++11 is a hard requirement for a while already.
Change-Id: Idb8fbdcd13398cc85fba583f40c2b5c4dc7c4943
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This update corrects several minor documentation errors that
cause qdoc warnings. These include incorrect or missing \fn
commands, incorrect uses of \e and \a commands together,
incorrect spellings, etc.
Change-Id: Ib26edef541fa3440025490bcf79cc101623e7f7b
Reviewed-by: Martin Smith <martin.smith@qt.io>
On macOS we pull out all the modifiers up front, which in the case
of incomplete key sequences such as "Meta+Shift+" will result in
an empty string as a result. The cross-platform code does not
handle that case, so we need to exit early.
This fixes an assert in tst_QKeySequence::parseString().
An assert has been added to QKeySequencePrivate::decodeString()
to make the assumption about the 'accel' argument explicit.
Change-Id: I135e62f9051a8b899202e5fb224b5d3c77bf2062
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The CLDR data contains eight locales with numeric territory codes, 001
for World, 150 for Europe and 419 for Latin America. The last was
already known in our enumdata.py, but as "Latin America and The
Caribbean", which is not supported by the CLDR, so I've amended it
while adding the other two. This gives us support for Esperanto and
Yiddish (among others).
[ChangeLog][QtCore][QLocale] Added support for World and Europe as
(numeric) "country" codes ("territory" in CLDR terms), thereby
enabling support for Yiddish and Esperanto, among other locales using
such codes.
Task-number: QTBUG-57802
Change-Id: Ibb1180fb720743a3a0589527649d10f3c9cd123d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
These tests kludge round Apple's use of GMT+1 and GMT+2 as names for
CET and CEST on Darwin; but 10.13 actually gets the names right, so
side-step out of the kludge when on this version or later.
Change-Id: Icb8a2f3ac30f0f621a19042dc03e0d281782dd41
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Erik Verbruggen <erik.verbruggen@qt.io>
It ignores its CollatorType, so we can use bool for it and save a few
bytes in QCollatorPrivate. The member using CollatorType follows three
existing bool members: a boolean will fit there, while an int requires
1-byte padding and its 4-byte payload.
Pointed out by Thiago Macieira.
Change-Id: I10c8ea6f1b735b1b872c509f18fd6a93e24c9b86
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QCollatorPrivate defines a back-end-specific CollatorType and sets a
member of this type to 0; but ICU's version really wants a nullptr not
0. So provide a named constant of the type, that's the NoCollator
value for use there.
Change-Id: Iad4d2f803ff4807ea568755efe00b9a92f1a8eeb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Both places that new'd it were then setting its locale themselves;
they might as well do it tidily by passing it to the constructor.
There's also no need to cleanup() in the constructor; every back-end
is a no-op when collator has its initial value.
Tidied up the class declaration in the process:
* moved {con,de}structor to be first methods
* comment on the two methods back-ends provide
Change-Id: I041669637935e68141e002156552af8b475ba36e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QCollator::detach() created a new QCollatorPrivate, then set some of
its fields to the values its constructor sets them to. Don't waste
time doing that.
Change-Id: I048a1553753ae2cf03d3f61d45490f2723ee658b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Each of its callers set dirty after calling detach() in any case, so
the call to init() was going to be duplicated (if the instance ever
got used again for something that benefits from it), and the d-ptr is
freshly created in any case, so can't be carrying any stray gunk from
earlier use, that init() might purge.
Change-Id: Id2485889c8ae4ee531c88562ff63f1a8ac0f4a36
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
They have a separate type that we can't recognize directly, but we
can check if we can recognize the QFlags<T> form, though we have to
add a lot of template-conditions to avoid triggering QFlags static
asserts.
Change-Id: I00853682c5376dd3cc411ff151f47bce2389e277
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Adds support for 16bit per color image formats in QImage. This makes it
possible to read and write 16bpc PNGs, and take full advantage of the
16bpc paint engine.
[ChangeLog][QtGui][QImage] QImage now supports 64bit image formats with
16 bits per color channel, compatible with 16bpc PNG or RGBA16 OpenGL
formats.
Task-number: QTBUG-45858
Change-Id: Icd28bd5868a6efcf65cb5bd56031d42941e04099
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Adds an enumName to QMetaEnum to carry the name of the enum since for
flags that doesn't match the name of the Qt type, but is needed if the
flag is scoped.
Change-Id: I1c0f77eb9e40e6fd1eb6a59bea77caf0f33fcf43
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Test that we don't silently replace an invalid TLS configuration with
the default one (for now, the only thing that is considered to be
non-valid - is having non-DTLS protocol set).
Change-Id: I6f714b009cf1345a085a3f26d638fc31330f1a94
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
This autotest fails on Ubuntu 18.04.
Task-number: QTBUG-69892
Change-Id: I3673d06f06fcd8db307fc53c27724b227978f20d
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
A weird behavior of the DTLS server example, when linked with 1.0.2,
exposed that client code, requesting an invalid protocol (for example, SSLv3)
can end-up with connection encrypted with DTLS 1.2 (which is not that bad,
but totally surprising). When we check the protocol version early in
setDtlsConfiguration() and find a wrong version, we leave our previous
configuration intact and we will use it later during the handshake.
This is wrong. So now we let our user set whatever wrong configuration they
have and later fail in TLS initialization, saying -
'Unsupported protocol, DTLS was expected'.
Auto-test was reduced - the follow-up patch will introduce a new
'invalidConfiguration' auto-test.
Change-Id: I9be054c6112eea11b7801a1595aaf1d34329e1d2
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
The one we had before was invalid and we ended up with
defaultDtlsConfiguration which has peerVerifyMode == AutoVerify.
Change-Id: I5b9ceb027e90189325c1d8fd0db37d1b212ebbc8
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The QTRY_VERIFY was needed because the window doesn't initially have a handle,
and QTest::qWaitForWindowActive(QWidget *) only checks the active state of the
widget if it does, returning false if not.
This broken logic should be fixed, but for now let's make it clear what's
actually going on by using an explicit wait for the window handle.
Change-Id: I6dd89e0894efed14f4b9a2562dfe8ca76b5ef89c
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
This updates the UI after the accent color or NSApp's effective
appearance have changed. For the accent color, we listen to
NSSystemColorsDidChangeNotification. For the effective appearance
changes, we do KVO on NSApp.effectiveAppearance.
Both changes will trigger rebuilding the system palettes followed
by a ThemeChangeEvent in the window system interface layer.
Task-number: QTBUG-68891
Change-Id: Iab1ec874e05f1f6d54cd60217c273e0f8ffbf49e
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
We were using CFRunLoopGetMain() everywhere. Get the
correct run loop using CFRunLoopGetCurrent() during
initialization, and store it.
Event dispatcher initialization must now be delayed
until after the constructor has run, since event dispatchers
may be created on the main thread and then moved to
the target thread. Initialize on first call to processEvents()
where the current thread will be the correct thread.
Use the stored m_runLoop instead of CFRunLoopGetMain().
This is required for wakeUp() and interrupt() which
may be called from another thread.
Change-Id: I6fffcfd4394899c4a12f241c42781979aaf99d5e
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
And reflow the text where it exceeds the 100 column limit.
Change-Id: I0d270c6a74a4c6ecba30e4e4d38a5d8f2cf81040
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
As discussed/proposed previously: remove the duplicated code when converting the
native certificate representation into QSslCertificate (configuration.peerCertificate).
Also, use the correct integer type when iterating - CFIndex is actually long, not int.
Change-Id: Ia6f43172e21b5153a93f1ef2589980d68ec2b39f
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
