When first starting an Android app we have invocation order issue, to
load the platform plugin we create the default QLocale (needed by the
resource locator code to see if :/qt/etc/qt.conf exists) so when the
android platform plugin loads and creates its own QSystemLocale, the
QLocale defaultLocalePrivate is already created and pointing to
globalLocaleData which means that systemData won't be called and thus
the code that triggers the call to QLocalePrivate::updateSystemPrivate
won't be called when calling QLocale().
I thought of two ways of fixing this, one was calling
QLocalePrivate::updateSystemPrivatea() from the QAndroidSystemLocale
constructor, but giving the responsibility to not break things to the
plugin seems a little fragile, so making the check on QLocale()
seems better.
Without this patch an Android app doing
QApplication app(argc, argv);
qDebug() << QLocale().name();
qDebug() << QLocale().name();
qDebug() << QLocale::system().name();
qDebug() << QLocale().name();
would print
""
""
"ca_ES"
"ca_ES"
now it correctly prints "ca_ES" the four times.
Task-number: QTBUG-41385
Change-Id: I2cf419f59aa008fa3aca11295fe7d42c40bcc32e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Silence debug output by setting QT_LOGGING_RULES
to turn off all debug output.
Task-number: QTQAINFRA-1631
Change-Id: I5c2366b4fe4bac341dcfd92f68b6da8071c5b089
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
Evaluating testlib's XML output in COIN would fail with:
"XML syntax error on line 7520: invalid UTF-8"
for the toLatin1() tests due to some Latin1/UTF8 mixup.
Add a helper function to convert the data to plain ASCII.
Task-number: QTQAINFRA-1797
Change-Id: I1c64878d4c2a67b8c2689905b5ffe6707b5963c1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The test has been crashing flakily recently.
1) It is creating a QGLWidget
2) It is stealing the QGLContext of that widget and moves it into a separate
thread.
3) In that secondary thread it makes the context current.
4) Meanwhile the QGLWidget itself may receive for example a resizeEvent or
other events and - since it assumes that it owns the context - attempts to
make it current.
5) Attempting to call makeCurrent() on a QGLContext that is in a different
thread than the current thread (via QObject thread affinity) will result
in a call to qFatal() and consequently the test aborts.
The conclusion from Simon Hausmann is that this test is testing a pattern from
Qt4 times that may or may not have worked back then. Nowadays with the Qt5
QOpenGL* API we do support this properly and there appears little sense testing
this.
Therefore remove the test altogether.
Task-number: QTBUG-66411
Task-number: QTBUG-66216
Change-Id: Ie2d66705bc7c3914ace6abcba9557c7c67ad4db3
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
This test executable was not flaky in the normal sense that when run, it
sometimes passes and sometimes fails. Instead, in some builds it would fail
consistently and in some builds it would pass consistently.
The first test to fail was version(ok00, default to last version) which gives
"mylib" as the library name and -1 as the library version. The description
implies that QLibrary selects the biggest or last used version when given -1.
However, versions less than 0 are not used at all. Instead the loading uses only
the name to select the library. Change the description to match.
So why did the test sometimes pass, sometimes fail? The test uses two library
projects lib and lib2 which install two different major versions of libmylib.
That includes the symbolic links:
libmylib.so -> libmylib.so.1.0.0*
libmylib.so.1 -> libmylib.so.1.0.0*
libmylib.so.1.0 -> libmylib.so.1.0.0*
libmylib.so.1.0.0*
libmylib.so -> libmylib.so.2.0.0*
libmylib.so.2 -> libmylib.so.2.0.0*
libmylib.so.2.0 -> libmylib.so.2.0.0*
libmylib.so.2.0.0*
The key thing being that both set the libmylib.so symbolic link. In a
multithreaded installation it's undefined which happens to set the link last.
The test code expected libmylib.so to point to libmylib.so.2.0.0. Ensure that by
building and installing lib2 after lib.
Task-number: QTBUG-66722
Task-number: QTBUG-66216
Change-Id: Ic513c772902273049c28e43fc1d83d550aafcd23
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
[ChangeLog][QtCore][QJsonValue] fromVariant() conversion now converts
from QUrl and QUuid using special encoding forms to ensure best JSON
compatibility.
Change-Id: I56b444f9d6274221a3b7fffd150cdc5ca1f87ff1
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
For compatibility with other parsers that may expect it to be so.
Change-Id: I56b444f9d6274221a3b7fffd150cd66390f98fd5
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
After a37785ec76 went in, it become
apparent that multi mouse button state handling in qtestlib is
non-existent, for details see QTBUG-64030 and QTBUG-63786. What
happened behind the scenes often was not what one would expect based
on the provided QTest::mouse* input sequence - events went missing,
incorrect events were generated, each subsequent test function
started with a state set from the function that run earlier. It is
easy to see how a minor change in one test could easily affect outcome
of other tests.
With a37785ec76, Qt platform plugins
are now responsible for sending explicit mouse button type and state
information; qtestlib should take full responsibility now as well.
But using the new API from a37785ec7 alone in qtestlib is not sufficient.
We need to reset mouse state between each new test function run (we do
this at function scope as that fits with the current qtestlib API user
expectations). This patch implements the necessary reseting logic.
Updated tst_qwindow.cpp::generatedMouseMove() to use QTest::mouse* APIs.
That test requires pressing multiple buttons, it was not possible with
QTest::mouse* APIs before this patch.
Added an auto test for multiple mouse button pressing/release in
tests/auto/testlib/selftests/mouse/. And few other tests which are
currently QSKIP-ed, but should be considered when re-designing qtestlib
APIs.
Task-number: QTBUG-64030
Change-Id: I39fdcbc73a467a7463ce2aed622bf22484095635
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Factor out a function to calculate the size and update the size when a
style sheet parent is set.
Task-number: QTBUG-64550
Change-Id: I3afe997085eae3cd48b7fe0a4c98582a8f572260
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Return value of qWaitForWindowActive should be checked as it is marked
Q_REQUIRED_RESULT.
One of the calls was failing, so remove it and wait for the next condition
instead.
Amends d0dffdfc01.
Change-Id: I0c0b2bf11504730477158a5e1fb3018f359a6fb9
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
The stdout of the processes used in the test was dumped if there was an error,
but the processes write their error messages to stderr.
Use MergedChannels process channel mode to dump both output streams.
Change-Id: I1645fd31c394da0871ee6ae36d37ca9a04d86052
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Before there was a risk looking up e.g index -1 if there
were no visible sections in layoutAboutToChange.
Change-Id: Ic911e4292e8e8c4892fef1c0f34cf7dccaad2bac
Task-number: QTBUG-65478
Reviewed-by: David Faure <david.faure@kdab.com>
The refactoring to support multiple providers broke alias resolving
(e.g. "text/directory" would be an invalid mimetype, instead of being
resolved to "text/vcard"). The unittest didn't catch it because most of
it was running with a single mime directory (and therefore a single provider,
in the new model). Fixed by re-running a number of test methods once we
have a second mime directory.
Change-Id: Ib5da89ba79c11ed41813b2aff4bc71c30afcde7d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The UBA in Qt was out of date, implementing the spec from pre
Unicode 6.3 days. It missed handling of directional isolates and
paired brackets.
This adds a completely new implementation of the UBA, that is
compliant with Unicode 10.
Added the test data from Unicode 10 to the qcomplextext auto
test and ensure that we pass the test suite.
Task-number: QTBUG-57743
Change-Id: Ie2d957bc9775d82f0a51d1c78dc6bd154f22847c
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Has been flaky in CI.
Task-number: QTBUG-66708
Task-number: QTBUG-66216
Change-Id: I69878574a98139681100c1424d5bbf46cc4a87b8
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
The test's client processes are prepared for the server not being ready when
they try to connect and handle QLocalSocket::ServerNotFoundError by waiting and
trying again.
However, on Ubuntu 16.04 and 17.10 and possibly other systems, sometimes the
error returned by qt_safe_connect inside QLocalSocket is ECONNREFUSED instead of
ENOENT. This has caused flaky failures in CI, so wait and try again in the case
of QLocalSocket::ConnectionRefusedError also.
Task-number: QTBUG-66679
Task-number: QTBUG-66216
Change-Id: I61e3d5b052d84c5ba9d1746f2c71db37cedbf925
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
Has been failing a lot lately
Task-number: QTBUG-66247
Change-Id: Id940a573eb299379cacceb836890cbe0b3c896b7
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
If _Thread_local is used on a block-scope declaration, it must be
combined with either static or extern to decide linkage.
Change-Id: I228b3520767197c6cdf5134ff5a666ab2aca33ea
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
They can be, if compiled with -Wl,-pie. Example:
$ file /usr/bin/ping
/usr/bin/ping: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 3.2.0
And you can't detect via the interpreter, since libraries can have
them too:
$ file /lib64/libc-2.26.so libQt5Core.so.5.11.0
/lib64/libc-2.26.so: ELF 64-bit LSB shared object, x86-64, version 1 (GNU/Linux), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2
libQt5Core.so.5.11.0: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 3.17.0
Change-Id: I940917d6763842499b18fffd15143bb80ce0e531
Reviewed-by: David Faure <david.faure@kdab.com>
Do not clear the settings of QHeaderView during layoutChange when the
model is empty and the section count did not change. This will not work
when a section is moved or a section is replaced with a new one during
layoutChange. But since layoutChanged is also called on sorting, this
patch ensures that the settings are not cleared in this case.
This restores the behavior to the same as before 5.9.4.
Task-number: QTBUG-66444
Task-number: QTBUG-65478
Change-Id: I39989cfd45b42e58f49d18ec014d3a941cadb6c9
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
During (re)storing the sections within layoutChanged handling, the
hidden section size was not properly stored which lead to a section
size of 0 when the section was unhided afterwards.
Task-number: QTBUG-66413
Task-number: QTBUG-65478
Change-Id: I0b714c7e0530a1eae82b3bb0e0dc80ed576522d0
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Currently there is a bug in Qt regarding the PostgreSQL driver as it
does not correctly escape the table names when constructing queries
internally. Therefore, these tests are marked as expected failures until
the bug itself is fixed in Qt.
Change-Id: I74dadc187f8a08509128dfea27be99787e57ea51
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This also accounts for some quirks on the PostgreSQL side:
- Null values for a related table are placed in a different
order when sorted.
- Functions (sum, count) return a different type than other databases
- Using quotes to account for case sensitivity with tables
Task-number: QTBUG-63861
Change-Id: Ib1894fa8d0c77d7045941f7c57be0d0acd8d117e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
TouchEvent::TouchPoint::pos was not updated in
QGraphicsScenePrivate::updateTouchPointsForItem().
To prevent the transformation being calculated repeatedly for each touch
point member, extract a function genericMapFromSceneTransform()
from genericMapFromScene() returning the transformation and use
that whereever multiple points are transformed.
Add a test, extracting helper functionality from
tst_QGraphicsItem::touchEventPropagation().
In addition, fold tst_QGraphicsScene::checkTouchPointsEllipseDiameters() from
c48f4bde00 into this test, so that
it is testing all transformations.
Task-number: QTBUG-66192
Change-Id: If71886d2c14c4e216f7781ea2f22f1adc444e6cf
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Since the PlatformInputContext is shared between test cases, the set commit
string from inputMethod() was otherwise saved and could pollute other tests.
For example on Windows PlatformInputContext::commit() was called when the
QLineEdit was first clicked onto in tst_QLineEdit::testQuickSelectionWithMouse()
and inserted "text" in the middle of the text, rather than starting selection as
intended.
This led to tst_QLineEdit::testQuickSelectionWithMouse() failing on Windows at
first in CI, but then always passing the repeats when it was tested alone.
Task-number: QTBUG-66499
Task-number: QTBUG-66216
Change-Id: Id6bdd263d57fd6d08fb48f013542b55b132907ea
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Fix 'Too many segments for object format' errors for (Debug) builds
using Windows ICC.
Change-Id: Ie48f43199948477c426d0a4e557f039eda129b22
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Fix 'unresolved external symbol __imp_Reg*' errors for builds
using Windows ICC.
Change-Id: I99cb6d53c45cadb31b5675182753f168a7bf4ea3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Switch to QTRY_COMPARE since thread-local references might be held shortly
after finished is signalled.
Change-Id: Ia32f1f45f6cc461352558e0f2acf9612f8a4639e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When the focus is lost on an editor due to the application no longer
being the active one then we have to ensure the parent view is going to
get the focus when it is returned. Since the editor does not have focus
when this check is done we need to manually account for this case by
setting it on the parent view as if it would if the editor did have
focus.
Task-number: QTBUG-62253
Change-Id: I14ac347e9e3a2bfaa8715a45811b17c1c7cf15f8
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Has been failing on it, but not on any other platform.
Task-number: QTBUG-66396
Task-number: QTBUG-66216
Change-Id: I0b208c675a23fb4bc1808dd3aa4dfef9bddf136b
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
Previous version of Ubuntu used to be blacklisted and the test is still flaky on
Ubuntu 16.04.
Task-number: QTBUG-66390
Task-number: QTBUG-66216
Change-Id: Iec404879f61164b995f0df7348f4f4baf608ca90
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
The test sometimes ended up with:
QThread: Destroyed while thread is still running
Received a fatal error.
This was because as a member variable of the local struct the QThread object was
sometimes destructed before the signal connection quitting it was handled. Fix
that by making sure that the thread is finished before finishing the test.
Also moved connecting to the state machine's signal to be before starting the
machine. Because the counting of QStateMachine::finished signal could hit 1
after the first signal is emitted and the test could pass without the code
working, check that both of the signals have been emitted.
Task-number: QTBUG-66372
Task-number: QTBUG-66216
Change-Id: If14141e39f37541032ddd8c6471daf40a77b0469
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
Without this fix, a fresh clean build of 5.9 will fail.
Change-Id: I69e4da382b07cc6e5e280e99478cbc3d44aa3f27
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
It is flaky on Ubuntu 16.04 and openSUSE 42.3.
Task-number: QTBUG-66345
Task-number: QTBUG-66216
Change-Id: I06fb88ee65113136309a0faa0336dd11672bfe59
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Remove BLACKLIST files which are no longer valid because the mentioned
CI systems are no longer active:
- opensuse-13.1
- opensuse-42.1
- rhel-7.1
- rhel-7.2
- rhel-7.3
- ubuntu-14.04
or the testcases are no longer available:
- QTBUG_14292_filesystem in qactiongroup
Change-Id: I80a4397059fafba169096440fdc07d45c76a1ed8
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
The possible key sequences for QKeySequence::Copy on X11 is Ctrl+C, then
Ctrl+Insert and at last F16. The order is defined in
QPlatformThemePrivate::keyBindings.
Task-number: QTBUG-46053
Change-Id: I86a0767e268088edfce98cfb07f9fb78f00d0713
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
If more than one VM tries to run the test at the same time, it times out. These
sharing violations were attempted to be worked around in 1c3dc8cfb, but the
workaround just leads to timeout, not success.
Task-number: QTQAINFRA-1727
Task-number: QTBUG-66216
Change-Id: If8bfd60dbb6575843680971d45b1c82e5beff534
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Likely fail is 3d29a0dd74. Is preventing
merge from dev, so blacklisted for now.
Task-number: QTBUG-66324
Change-Id: Ieb522897e93d4a1b4c0e423fadd4138dfb089093
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Conflicts:
src/corelib/tools/qvarlengtharray.qdoc
src/corelib/tools/qvector.qdoc
Resolved documentation changes in favor of 017569f702,
which keeps the move overloads along with its const-ref sibling.
Change-Id: I0835b0b3211a418e5e50defc4cf315f0964fab79
This test was blacklisted in 5.10 on opensuse, but it fails a lot on
ubuntu as well. While we are discussing a real fix for this, temporarily get this out of
the way.
Task-number: QTBUG-63152
Change-Id: I4f1d3b261013052636ee13eda30f94b647a43a38
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The peer Tests for QDBusAbstractAdaptor are so flaky that it's
very difficult to get any commits through to qtbase.
Task-number: QTBUG-66216
Change-Id: I8da80f71aa832e683f72129cb2d4785425d39c00
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
When calling setSceneRect() on a QTouchPoint it will cause the
ellipseDiameters to be changed, whereas this should not be affected by
the scene rectangle as it should be in logical pixels.
Also add a manual test for visually checking the ellipse diameters on
various devices.
Change-Id: I1ee9207cb1a63cfef33fe904594c73aba221af5c
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Timeouts with subsequent failures to delete the temporary
directories have been observed in COIN.
Previously, QProcess:terminate() was used to end the processes,
which does not have any effect on console processes on Windows.
Add a helper function which resorts to kill() on failure
to terminate().
Change-Id: I05539d1703280d34b392f2e8ff8565b9a04d703c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
In order to save having to always run a query to get the tablename for
a known oid then we cache the result on the driver side. The oid stays
the same while the table exists, so only on dropping it would it change.
Recreating the table causes it to get a new oid, so there is no risk of
the old one being associated with the wrong table when this happens, if
the driver is still open at that point.
The benchmark added shows the improvement from the previous code, before
the results for PostgreSQL was:
RESULT : tst_QSqlRecord::benchmarkRecord():"0_QPSQL@localhost":
259 msecs per iteration (total: 259, iterations: 1)
whereas now it is:
RESULT : tst_QSqlRecord::benchmarkRecord():"0_QPSQL@localhost":
0.000014 msecs per iteration (total: 59, iterations: 4194304)
Task-number: QTBUG-65226
Change-Id: Ic290cff719102743da84e2044cd23e540f20c96c
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
Reviewed-by: Robert Szefner <robertsz27@interia.pl>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
QFontEngine::cloneWithSize() is used by QRawFont internally when switching
a raw-font from one size to another using setPixelSize. For CoreText, we
use a subclass of QCoreTextFontEngine to keep track of the QByteArray data
of a raw-font, but failed to overload cloneWithSize, so we would lose the
data whenever setPixelSize was called, resulting in missing text rendering
in QtWebKit. We now retain the data as we should.
Task-number: QTBUG-65923
Change-Id: I7d4186a3c32a61d48d1e9388e43f2792e8e46081
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
This way, it's lossless.
This commit is a cherry-pick of ab1e507574,
which was reverted.
[ChangeLog][QtCore][QVariant] Conversions of QDateTime to strings now
contain the millisecond components.
Change-Id: I5e421e32396d44e4b39efffd150b744e40fff3a1
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@qt.io>
Commit 8f52ad9fe0 ("ucstricmp: compare
null and empty strings equal") made sure empties and nulls would compare
equally, but may have broken the null vs non-empty comparison (which was
not tested). The commit message also said that it expected all callers
to handle null before calling into those functions, but that's not the
case for QStringView created from a null QString: the incoming "a"
pointer was null.
So just remove the checks for null pointers and rely on the size checks
doing the right thing.
[ChangeLog][QtCore][QString] Fixed a regression from 5.9 that caused
comparing default-constructed QStrings to be sorted after non-empty
strings.
Task-number: QTBUG-65939
Change-Id: I56b444f9d6274221a3b7fffd150c83ad46c599b6
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
While we're at it, add a way to get it without the dashes too. I'm
calling it "id128", as in "128-bit ID", as seen in journald's sd_id128_t
type and the sd_id128_xxx() API.
[ChangeLog][QtCore][QUuid] Added a parameter to both toString() and
toByteArray() to allow controlling the use or not of the braces and
dashes in the string form.
Change-Id: I56b444f9d6274221a3b7fffd150cde706cfc5098
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
This is a standard feature in GtkEntry widgets or HTML
<input type="text"> elements. During a normal text selection by mouse
(LeftButton press + mouse move event), it's now possible to quickly
select all the text from the start of the selection to the end
of the line edit by moving the mouse cursor down.
By moving it up instead, all the text up to the start of the line edit
gets selected. If the layout direction is right-to-left, the semantic of
the mouse movement is inverted.
This feature is only enabled if the y() of the mouse move event is
bigger than a fixed threshold, to avoid unexpected selections in the
normal case. This threshold is set by the QPlatformTheme and a value
smaller than zero disables this feature.
The threshold is updated whenever the style or the screen changes.
[ChangeLog][QtWidgets][QLineEdit] Implemented quick text selection by
mouse in QLineEdit.
Change-Id: I4de33c2d11c033ec295de2b2ea81adf786324f4b
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
This reverts commit ab1e507574. That was
supposed to be a minor behavior change, but ends up having visible
effects such as QtXmlPatterns xs:dateTime type now reporting sub-second
fractions. So we're reverting in 5.10 and re-applying in 5.11.
Change-Id: I741e49459c9a688c1c329d6cbd521cd4a0b2aa84
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Actually check that there's a T where ISO 8601 wants it (instead of
just skipping over whatever's there), with something after it; move
some declarations later; add some comments; and use the QStringRef API
more cleanly (so that it's easier to see what's going on). Simplify a
loop condition to avoid the need for a post-loop fix-up.
This incidentally prevents an assertion failure (which brought the
mess to my attention) parsing a short string as an ISO date-time; if
there's a T with nothing after it, we won't try to read at index -1 in
the following text. (The actual fail seen had a Z where the T should
have been, with nothing after it.)
Add tests for invalid ISOdate cases that triggered the assertion.
Task-number: QTBUG-66076
Change-Id: Ided9adf62a56d98f144bdf91b40f918e22bd82cd
Reviewed-by: Israel Lins Albuquerque <israelins85@yahoo.com.br>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
(cherry picked from commit a9c111ed8c)
- Allocate widgets on stack. The previous version was inconsistent (some
widets were managed by QScopedPointer and some were simply leaking).
- Use QTest::mouseMove(QWindow *)
- "This test fails on OS X on CI" was a clear indication that test is
flaky, the new implementation can reliably reproduce issue which is
now tracked in QTBUG-63031.
Task-number: QTBUG-63031
Change-Id: I59965ef8fa8edca17c8a73901d81e9efc7da3c5b
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
QHeaderView is doing a complete rebuild of the sections when the layout
changed because everything could have happened. But since layoutChanged
is also called during e.g. sorting, the old data must be restored when
possible.
Task-number: QTBUG-65478
Change-Id: I088d4d843cad362b97df6dc5e0dcb9819b13547f
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
This is in preparation to adding CBOR support. We don't need yet another
dir for CBOR and placing it in src/corelib/json is just wrong.
Change-Id: I9741f017961b410c910dfffd14ffb9d870340fa6
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
QAbstractItemModel::dataChanged() gained an optional role parameter
with Qt5 which was not filled within QListWidgetItem/QStandardItem
setData() functions
Task-number: QTBUG-55903
Task-number: QTBUG-63766
Change-Id: I4da9346ef8401cc8633dc4b2ea7d00451d1e3942
Reviewed-by: Luca Beldi <v.ronin@yahoo.it>
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Wait a little bit longer for the visibility and resize events. By
directly waiting for the incoming events with QTRY_COMPARE we give
the test some more time to deliver the expected events.
Also move the check for the expectedResizeCount/expectedPaintCount
before the geometry check for better debugging.
Task-number: QTBUG-64639
Change-Id: I712e432aa8f8d35dd199adf88af3653009528b57
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
File system operations like renaming/removing may fail on Windows when
file system watchers are present. Add functions to
QFileSystemModelPrivate to temporarily remove the watchers prior to such
operations and to restore them in case of failure. Use them for
rename/remove (within a feature check for QFileSystemWatcher
and Q_OS_WIN).
Task-number: QTBUG-65683
Change-Id: I90142901892fbf9b1e1206a3397a95ffd3c8f010
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
With this change it is possible to use all supported
configurations in different backends without any new interfaces.
Change-Id: Ib233539a970681d30ae3907258730e491f8d3531
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Instead of having to modify the tst_databases.h file whenever you want
to add a test database, this will now read from a file to find what
databases are available. This defaults to dbs.json in the same directory
or the QT_TEST_DATABASES_FILE environment variable can be set to point
to the file that contains the databases. The latter makes it easier for
CI then to have something set up on a per configuration basis.
The SQLite database stays hardcoded so this will continue to be tested as
before without any additional changes.
Change-Id: I7e7ccde6a6be3e490dd640b1590e3b691a6b2ab3
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
The include is needed for std::unique_ptr on winrt.
Change-Id: I72a28bd0951cc947ac65877ccc35f464c757c444
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@qt.io>
The backwards iteration was done under the assumption that the only
valid modification of the winEventNotifierList in a slot connected to
activated() would be the removal of the notifier itself. This is wrong.
Instead, iterate forwards, like before 85403d0a, and check the index
against the current list size in every iteration. This ensures that we
do not run out of bounds while the list is modified.
Also, retry the activation loop if the list was modified by a slot
connected to activated(). This ensures that all notifiers with signaled
handles are activated.
Task-number: QTBUG-65940
Change-Id: I25f305463b9234f391abc51fe0628d02f49b6931
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
This way, it's lossless.
Change-Id: I5e421e32396d44e4b39efffd150b744e40fff3a1
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The idx member of EventWithNotifier is unused.
Change-Id: I0f5aacaaad4b4e82c57ff7bb020586944014f139
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Adds inline handling of the C++17 type std::variant, so the type will
be resolved if converted into a QVariant.
Change-Id: I31809d70d7f347277389d42a3695836ec7a32d02
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Commit 29bc68cf16 added support for
unsigned and commit 5ff7a3d96e later added
support for int. This commit adds support for qsizetype, which isn't int
on 64-bit platforms.
We do this by reorganizing the code and using the generic version of
__builtin_{add,sub,mul}_overflow from GCC 5 and Clang 3.8, which ICC 18
seems to support now too on Linux. That leaves older versions of GCC and
Clang, as well as MSVC, ICC on Windows, and the GHS compiler, to use the
generic implementations, as I've removed the assembly code those
versions of GCC and Clang on x86 are now uncommon.
Note: any older version of ICC probably breaks. We only support the
latest.
Change-Id: I9e2892cb6c374e93bcb7fffd14fc11bcd5f067a7
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Use qInfo() instead of qDebug() in case QDebug is turned off.
If the 'diff' tool is available, write the output to temporary files and
run diff on it. Otherwise, print the lines as was before, but onto one
stream to avoid indentation by the testlib handler.
Change-Id: Ib5a5dfb66ce481b493b85b915aa8c785ecb6b387
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
While menubar actions are hidden, as menubar was too narrow
to contain them, by clicking the blank area, where the action
was supposed to place, can still trigger the action.
Task-number: QTBUG-65488
Change-Id: I6b137e0717f634ebd3371dbcc2c1ce2089688374
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Setting speed to 0 means the current frame will continue to be shown,
the finished signal is not emitted, and state remains QMovie::Running.
Task-number: QTBUG-65758
Change-Id: I681d902e3211c5899b21043e5177b7c73d5d3fb5
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Passing -1 to waitForReadyRead() may cause it to wait for some time
but the data retrieved may be enough for processing. So if 0 is passed
from read, indicating that there is potentially more to come, then
it will do a waitForReadyRead() then for more data to come.
Change-Id: I75f270d1f124ecc12b18512cc20fb11f7a88f02e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
By ensuring that the current index follows the focus of an index widget,
we can ensure that moving the cursor will happen in the way that is
expected from the focused widget.
Task-number: QTBUG-27793
Change-Id: Ia36891a94ce41c7d12fba678de23a6f3b69374ae
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
The action which added by QMenuBar::addAction(const QString &text) already
connected relevant signals and slots implicitly, however, while
QMenuBarPrivate::updateGeometry -ing, it reconnects them
if there's a extension button associated with a hidden popup menu.
In that case the QMenuBar::triggered would be fired twice.
Since the QAction's ownership may be changed or added dynamically,
there are still very rare cases like several widgets share the same
QAction object to result in this problem.
[ChangeLog][QtWidgets][QMenu] Fixed a bug in QMenu that caused QMenuBar::triggered
to be fired multiple times.
Task-number: QTBUG-25669
Change-Id: I4d52e82a2136a992e0b37118e41237d96a2c5d22
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
The Embedded Android build (Boot to Qt Android injection) is defined by
having both Q_OS_ANDROID and Q_OS_ANDROID_EMBEDDED flags defined,
as well as having Qt config android-embedded.
This commit enables the possibility to build embedded Android builds.
(i.e. Qt build for Android baselayer only, without JNI)
Change-Id: I8406e959fdf1c8d9efebbbe53f1a391fa25f336a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
The size of the QStaticText was always adjusted, even if setTextWidth()
was used. Now size of the QStaticText is calculated according to
the set width of the text, and if no width was set, then the
automatically adjusted size is used.
[ChangeLog][QtGui][QStaticText] Fixed explicitly set width not being
respected.
Task-number: QTBUG-65836
Change-Id: If2f9f6952fb168f4bcb6d8fabfdc7360f8a36485
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Improves performance and STL compatibility by adding rvalue versions
of prepend and insert.
[ChangeLog][QtCore][QVarLengthArray] Added rvalue overloads of
prepend and insert.
[ChangeLog][QtCore][QVector] Added rvalue overloads of prepend
and insert.
[ChangeLog][QtCore][QVarLengthArray] Can now contain movable but
non-copyable types, such as std::unique_ptr.
Change-Id: I6c946acc5b67502c91c52ac5dea67cedb1af93a5
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is a regression test for bugs resulting from failure to take
account of the MS TZ APIs fake-DST handling of standard-time
transitions happening during DST. Has to skip the 2014 test on Win7,
as it's so old it doesn't know about the (for it) "future" transition.
Task-number: QTBUG-42021
Change-Id: I853b86d03a86f0269371bf1622bf63882fb8ee5d
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QItemDelegate::doLayout added a margin after the checkbox which is not
done within QCommonStylePrivate::viewItemLayout. This created a small
but visual difference between using QItemDelegate or QStyledItemDelegate.
The additional margin was removed from viewItemLayout somehwere between
4.4 and 4.5 so it was intentional.
Change-Id: I2fc3f287fe8b181355edeebec9626c49d85fe74d
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
For Q_OS_WIN, a path is only truly absolute if it includes a drive
letter; merely starting with a slash is not enough. (We can't support
UNC paths, so don't even try: qmake runs various commands in the
source directory using CMD.exe, which doesn't support UNC as PWD.)
This requires, when resolving a path relative to a root, transcribing
the root's drive to such not-quite-absolute paths.
Changed QMakeGlobals, $$absolute_path() and $$relative_path() to now
use IoUtils::resolvePath() rather than delegating to QDir's absolute
path method, since that doesn't correctly recognize the need for a
drive letter (and qmake did run into problems with some paths, from
splitPathList and a failing test, as a result).
Moved existing ioUtils tests for handling of relative / absolute paths
out into separate functions and expanded significantly. Fixed some
existing tests to use an absolute path where one is needed; added two
tests involving driveless (but rooted) paths; and fixed the test init
to set a value for QT_HOST_DATA/src property (the lack of which lead
to an assertion failure with this fix).
Task-number: QTBUG-50839
Change-Id: I2bfc13c1bfbe1ae09997274622ea55cb3de31b43
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Use Ids from newly introduced id attribute depending
on the global form setting.
Change-Id: I0a5094d5543c0714c88511fa159b60afc9be3c81
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
When QFile::open is called with the NewOnly flag, the call will
fail if the file already exists. As usual, if the file does not exist,
it will be created. Like QTemporaryFile, there is a guarantee from
the operating system that you are not accidentally creating a new file
on top of an older file. When QFile::open is called with the
ExistingOnly flag, the call will fail if the file does not exist. The
ExistingOnly flag only provides new functionality when used with the
WriteOnly flag. For ReadOnly it provides no change in functionality,
as ReadOnly by itself already never creates.
Task-number: QTBUG-52244
Change-Id: I8e3206728f245f95172c225bf297023fb078fc6d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Generate it only when the form contains some actions.
Change-Id: Ic1d64003ccff6102174771c04999cf7c90ac2ae8
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Generate it only when needed.
Change-Id: I7a89b18ec0f2ee7d55fcad1b3f9701269d7616ec
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
... that will be used if an icon can't be found in the
current theme.
The Icon Theme Specification
https://standards.freedesktop.org/icon-theme-spec/latest/ar01s05.html
states that unthemed icons must be searched in the base directories,
i.e. /usr/share/icons, ... But in practice unthemed icons are
installed into /usr/share/pixmaps and this dir is not used as
a base dir for icon themes. So it's better to explicitly specify
fallback dirs to avoid needless access to the filesystem.
Also some KDE application install their own unthemed icons
(into /usr/share/<appname>/pics), that can't be found by
QIconLoader. With this change it would be possible for them
to specify dirs with unthemed icons and thus be displayed
correctly in non-KDE environments.
[ChangeLog][QtGui][QIcon] Added fallbackSearchPaths() that
will be used to find icons missing in the current icon theme.
Change-Id: I0dc55ba958b29356a3b0a2123d6b8faa24d4c91e
Task-number: QTBUG-33123
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Data in item models is most often organized in rows, where each column
contains an attribute of the item represented by the row. Often when
sibling is used, it is to request another piece of data from the same
row. Having a specialized version makes this easier and less awkward
to do, simplifying
auto sibling = index.sibling(index.row(), columnOfInterest);
to
auto sibling = index.siblingAtColumn(columnOfInterest);
For symmetry reasons, siblingAtRow(rowOfInterest) was also added.
Change-Id: Ib203b2cdb16154cbb2680d16fb5c6a7538f33d07
Reviewed-by: Topi Reiniö <topi.reinio@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
Apparently this (undocumented) TESTDATA feature creates resource files,
but lets you use wildcards as well, which is very handy.
The reason I didn't know/realize this when adapting the tests to use a
".qrc"-file* was because some of the test-cases were using relative
paths instead of the 'testDataDir' variable.
This commit fixes the remaining uses of relative paths, removes a
usage of QDir::setCurrent, and adapts QSslSocket to use TESTDATA.
* in now-reverted commit e1600c1a73
Change-Id: Iee6d88f1e0810eeaadac90e7d44bc6db84bfeabf
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
The operator_eqeq(data13) test expected the local-time epoch and UTC
epoch to agree precisely if the localTimeType set by the test's
constructor says local time is UTC; however, when the local zone is
*sometimes* ahead of (or behind) UTC, due to DST, localTimeType is
duly set to indicate that, which doesn't preclude the zone agreeing
with UTC at the epoch. This indeed happens for Europe/London, which
agrees on the epoch but was ahead a few months later. So we can't
determine what outcome to expect based solely on localTimeType,
although we can be sure of a match when local time is UTC. So skip
this test when local time isn't UTC (and document what's going on a
bit better).
Task-number: QTBUG-65435
Change-Id: Id9b8aa0402f2a2b410e0234f6eca4ab0d1010bc4
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Remove usage of outdated header.LGPL21 and replace those with proper
one (header.LGPL in src, header.GPL-EXCEPT in tests)
Change-Id: Ia4d1c0d84b77f09787fe7c30670747a1fe2aff29
Reviewed-by: Liang Qi <liang.qi@qt.io>
This removes the need to install an event filter on qApp just for this.
A similar thing was done with setPalette to reduce the number of event filters
in e.g. SystemPalette and Quick Controls.
[ChangeLog][QtGui][QGuiApplication] Added fontChanged signal
Change-Id: Ifa843aa42b91ac63ab17c3b064ac0e764aac77d3
Reviewed-by: J-P Nurmi <jpnurmi@qt.io>
Move the modeltest autotest in the right place, and fix the
other autotests that were using it to use the version now in QtTestLib.
Change-Id: Ic6838945f616d580f357c872ce0956c341be3b16
Reviewed-by: David Faure <david.faure@kdab.com>
We have the ability to blacklist tests for CI runs now.
Task-number: QTBUG-35109
Change-Id: I8590e83faba764dce2d52e8c62e2e2c63f7bf219
Reviewed-by: Liang Qi <liang.qi@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The code used to fall back to anonymous login independently for username
and password; however, it should only use a fall-back password if the
username is missing or (case-insensitive) "anonymous". When a
non-anonymous username is given without password, we should simply skip
he PASS message to FTP.
If the FTP server requests a password, in the latter case, QFtp will
signal authenticationRequired; in all cases, if the server rejects the
given credentials, QFtp signals authenticationFailed. Either way, the
client code can then query the user for credentials as usual.
Task-number: QTBUG-25033
Change-Id: I2a4a3b2725819ab19c8a7e4baa431af539edcd8d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
After we stopped sanitizing the fallback font list (with change
6ca48a847a), we now need to make
sure it is ordered so that the fonts that support the writing
system in question are always tested first, otherwise we can end up
loading a lot of fonts that will never be used.
Task-number: QTBUG-65605
Change-Id: Id2a65bbff3e64e6d6e6b4f72500778ee3e811e84
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Introduce a special hash modeled on the one used for QFileSystemWatcher
on Windows.
Task-number: QTBUG-31103
Task-number: QTBUG-64147
Change-Id: I69ebabe841716e4957ae3fb04fa5c43d233a3552
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Then we don't need to add a leading slash. (minor clean-up)
Change-Id: I86af224841009fda838e7cb89d47d324963328c9
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This reverts commit e1600c1a73.
The commit missed the fact that TESTDATA exists. Which supports
wildcards and then can automatically pick up new files when added (as
long as they match a wildcard) and then you don't need to maintain a
giant qrc file.
Change-Id: Ie31fadb5ef6e8dfe6105f4f9764292f78cffb512
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Old header.LGPL21 header was used at some files. Replase those with
new header.LGPL one
Remove old header.LGPL21
Task-number: QTBUG-57147
Change-Id: I650e39024ed4876bba27e954c7d61fdb025b46ef
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
When we request fallback fonts, we cannot discriminate the fonts
based on the writing system support. This is especially important
since common script is now merged with other scripts, meaning that
a common script character will always go through the fallback
mechanism when not supported by the main font. When drawing
for instance a string of Devanagari characters on macOS, we would
get a list of 33 fallback fonts, but almost all of them would be
the default Devanagari font, since none of the other fallbacks
would support that script. Meaning that we would just check the
same font over and over, which makes no sense. The fallback list
has been retrieved specifically for the given script, so we do
not need to consider that when fetching the fonts.
For most of the common set, we will not have noticed the bug,
because at least one of the writing system-specific fallbacks will
have had support for latin characters as well. But when trying to
mix emojis and some non-common script, we would get a box in
place of the emoji, which had been adopted to the main script and
would only be looked for in the fonts supporting this.
Note that this exposed an issue with the QRawFont test on some
systems. When the sample text contained a space, it would
be possible to get a fallback font for this character, since
we now effectively support fallbacks. This is not the correct
behavior, but it is unrelated to this fix, and it was not what
the QRawFont::unsupportedWritingSystem() test was written to
check. I have therefore removed the space from the sample text
to make the test pass, and will make a separate task of fixing
the issue of merging fonts for whitespace characters.
[ChangeLog][QtGui][Text] Fixed a bug where mixing different
writing systems with emojis could lead to missing glyphs.
Task-number: QTBUG-61882
Change-Id: I00f6043bb01af1f2277723ccf643034aebf3e18f
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
sqlite will reuse the index for bound parameters when the named
placeholder is duplicated so we only need to call bind one time for each
placeholder. Therefore we need to have just one instance of each value
when doing the bind.
Task-number: QTBUG-65150
Change-Id: I75c4bcc4563e43c180a59a7a4cbb770dbe994642
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
When implementing a custom model there's the habit, in each and every
function that takes a QModelIndex, to carefully checking the index
passed by the caller. This index is checked for "legality" (*): does the
index belong to this model, is the index pointing to an existing row and
column, and so on. These checks are hand-rolled and, as such, slightly
different and possibly incomplete (i.e. wrong) every time.
What's worse, these checks are implemented via "ordinary" code (if
statements). However, passing an illegal index to a QAIM function is a
precondition violation, and as such does not (and must not) be
checked in ordinary conditions, as it triggers undefined behavior. On
the other hand, while debugging a custom model or a custom hierarchy
of (proxy) models, having such checks in place can be a significant
aid.
Enter checkIndex(): a debugging helper for QAbstractItemModel and its
subclasses. checkIndex() centralizes the checks for legality of a
given index. User code is free to assert on it, or have some other
fallback mechanism in case a check fails.
(*) Using "legality" here instead of "validity" in order to avoid
confusion between QModelIndex::isValid() and what checkIndex() really
does.
[ChangeLog][QtCore][QAbstractItemModel] Added
QAbstractItemModel::checkIndex(), a debugging function for
QAbstractItemModel subclasses.
Change-Id: I1eea0586b1ac3ededdbfbf46759145022dc5ad86
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Reviewed-by: David Faure <david.faure@kdab.com>
This change adds support for multiple result sets in PostgreSQL.
[Important Behavior Changes] The QPSQL driver now supports multiple
result sets. Since QPSQL previously did not support multiple result
sets, there may be some compatibility issues with the existing code
that executed several queries as one and were expecting to get
the results of the last one. In this case use QSqlQuery::nextResult()
to move to the last result set.
[ChangeLog][QtSql][QPSQL] Added support for multiple result sets
Change-Id: I2bfc91f512c4dac83116f3aa42833839a6da084c
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Add some macros to enable systematic testing of QStringBuilder
expressions.
Change-Id: I6b7dbcb4a4ca0d1aebdfe7cb9861af881c0cd346
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Remove black-listing of test; the problem was due to misconfigured
time-zone on the CI system - it was on some zone that presently
coincides with UTC, not actually on UTC as supposed.
This reverts commit 597b96b8fa.
Change-Id: I72ad1dfa38532362c05aef33cd874f7f79879a41
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
[ChangeLog][QtWidgets][QMenuBar] Add overloads of addAction() using Qt 5 signals and slots
Change-Id: Ief21974213b80111f0ca87df490eb72dd6b9c9b9
Reviewed-by: Martin Smith <martin.smith@qt.io>
With this change, it is possible to significantly reduce memory
consumption of applications that fetch large result sets from databases.
The implementation is based on the new functionality called "single-row
mode" that was introduced in PostgreSQL version 9.2:
https://www.postgresql.org/docs/9.2/static/libpq-async.html
It also uses asynchronous commands PQsendQuery(), PQgetResult():
https://www.postgresql.org/docs/9.2/static/libpq-single-row-mode.html
[ChangeLog][QtSql][QPSQL] Added support for forward-only queries (requires
libpq version 9.2 or later)
[Important Behavior Changes] The QPSQL driver now supports forward-only
queries. To use this feature, you must build QPSQL plugin with PostreSQL
client library version 9.2 or later. See the Qt SQL documentation for
more information about QPSQL limitations of forward-only queries
(sql-driver.html).
[Important Behavior Changes] If you build the QPSQL plugin with PostgreSQL
version 9.2 or later, then you must distribute your application with
libpq version 9.2 or later. Otherwise, the QPSQL plugin will fail to load.
Task-number: QTBUG-63714
Change-Id: I15db8c8fd664f2a1f719329f5d113511fa69010c
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
QHeaderView::resizeSection() did not check if the given section size is
inside the min/max property bounds.
Also on calling setMin/MaximumSectionSize() the current section sizes
were not checked if they are inside the new given bounds.
This is a small behavior change when a user is setting the section size
via resizeSection() without respecting the min/maxSectionSizes.
Task-number: QTBUG-64173
Change-Id: Ia9c9eebf058d60c776ab5f8f8336642013ec553f
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Since there may be some applications that want to make their comboboxes
scrollable or not with the mouse wheel then a style hint is added to
make this toggable. It defaults to true for all platforms except macOS
to keep it in line with existing behavior.
[ChangeLog][QtWidgets][QStyle] Added SH_ComboBox_AllowWheelScrolling as
a style hint to enable/disable the use of the mouse wheel in a QComboBox.
This defaults to true in all styles except the macOS one so there is no
change in existing behavior.
Task-number: QTBUG-53094
Change-Id: I6eb61e1990bfc79e3b3ea82d7b91ee39096fa077
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
Setting the same brush on the same group and role should
not detach nor alter the result of QPalette::isCopyOf().
Task-number: QTBUG-56743
Change-Id: Ic2d0dd757d703b01e8c5d835a8c124b3317653f4
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Files to test binary compatibility for 5.10 added. It looks like
tst_bic (that runs the BC tests) is implemented so that it will start
testing new content when e.g in case of 5.10 branching to 5.10.1 has
happened.
Change-Id: Ic97db0339cc2942c088962dda495eb42b96fe153
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QTreeWidget::visualItemRect() returned an invalid QRect when a column
was moved and then hidden (or the other way round). The reason was that
the logical index returned by QHeaderView::logicalIndexAt() was again
passed to QHeaderView::logicalIndex() to create the QModelIndexes needed
for QTreeView::visualRect()
Task-number: QTBUG-28733
Change-Id: I8676f21bcab8c05c2260b85d483902f18cbf3e24
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Also adjusted the text segmentation and line break algorithms
so that they can handle the new data, and pass the test suite.
Change-Id: Ib727fd80003e34e96458d7a681996de3fa3691e7
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Make sure we properly handle both directional embeddings as well as
directional isolates to determine the direction of the string.
According to the latest version of the Unicode bidi algorithm,
parts of the string contained inside an directional isolate is
to be ignored when determining the paragraph direction. Embedding
markers themselves are to be ignored as well, but not the characters
inside an explicit directional embedding or override.
This is also some required pre-work to get our BiDi algorithm
updated to the latest version of the standard.
Move the implementation to QStringView and implement the methods
in QString and QStringRef through that implementation.
Task-number: QTBUG-57743
Change-Id: I7f24e09198e22d6359c6534c9ae40a904e94c46e
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
QTreeView makes columns movable except for the first one
(since Qt 5.0, see commit e0fc088c0c). This is indeed best
for actual trees, but not when using QTreeView as a flat
multi-column list (with no "root decoration"). Then it should
be possible to move the first column.
[ChangeLog][QtWidgets][QHeaderView] Flat treeviews can now allow
the user to move the first column (like in Qt 4.x) using the new
method QHeaderView::setFirstSectionMovable().
Change-Id: I6b5025e40850bf5c4c373124ee81f657f3f09d29
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
QSqlDatabase objects can only be used in the thread that the connection
was opened for. So if the driver was created already then we check if
the thread is correct. If it is not then we output a warning and return
an invalid QSqlDatabase.
[ChangeLog][QtSql][QSqlDatabase] QSqlDatabase::database() will return
an invalid QSqlDatabase if the calling thread does not own the requested
QSqlDatabase.
Task-number: QTBUG-216
Change-Id: Ib5a25aa62129e3925f9819109af05961e5178bc5
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This patch updates the code from QSharedMemory to remove the use
of the deprecated QRegExp class. It also updates the unique key
test to avoid change of behavior going undetected.
Change-Id: I649e615027507898800bb5454a62a6cf8bbb2e18
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Since its initial implementation, QRingBuffer had the following
fragilities in the architecture:
- it does not guarantee validity of the pointers, if new data will
be appended. As an example, passing an address of the QRingBuffer
chunk as a parameter to the WriteFileEx() function on Windows
requires the stability of the pointer. So, we can't add new data
to the QRingBuffer until the overlapped operation completed
(related issues were fixed for QWindowsPipeWriter and QSerialPort
in 5.6 branch by introducing an intermediate byte array);
- inefficient reallocations in reserve(), if a shared chunk was
inserted in the queue (we can get a reallocation in the place
where we don't expect it:
char *writePtr = buffers.last().data() + tail; <- line #133
).
Proposed solution is to avoid reallocation by allocating a new
block instead. That was accomplished by introducing a QRingChunk
class which operates on a fixed byte array and implements head/tail
pointers strategy for each individual buffer in the queue. So,
QRingBuffer is no longer dependent on QByteArray's internal
shrink/growth algorithms.
Change-Id: I05abab0ad78e22e4815a196037dfc6eff85325d1
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Unlike higher scopes (like scope 4, admin-local, which the last commit
used), scopes 1 and 2 require a scope in order to bind, even if some
operating systems are lenient. So test that we are able to bind to them
and do bind properly.
Change-Id: Ifb5969bf206e4cd7b14efffd14fba153eab965b9
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Binding without an interface and expecting the OS to select something is
not supported in all OSes. On FreeBSD, I keep getting EADDRNOTAVAIL. So
modify our test to only join, leave and send to multicast groups with an
interface selection.
With this, all tests either pass or are skipped for me on Linux,
FreeBSD, and macOS. On Windows, this revealed an inconsistency in
behavior, which this commit adds a workaround for.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb6815456494d2
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
tst_QUdpSocket::broadcasting and tst_QUdpSocket::pendingDatagramSize
fail on the new Ubuntu 16.04 clean template.
Task-number: QTBUG-65440
Change-Id: I0e973b9c90b7c5827406bac8138370b61992a115
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
This autotest fails on the new Ubuntu 16.04 template with UTC timezone
in the system settings.
Task-number: QTBUG-65435
Change-Id: I397f01ab3fed354a4eeec8b05415226a75fce5a1
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
The FreeBSD kernel treats them specially, just like link-local (that's
probably why it calls them "interface-local" instead of "node-local").
So instead let's use a random address, which will avoid multiple
tst_qudpsocket, when run on the same network at the same time,
receiving each other's datagrams. It could happen, considering this test
has an 800-second timeout limit.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb592a3166547e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
This is not an official feature of the networking stacks and does not
work portably across operating systems. So just stop trying to do that.
This was failing reliably (not flaky!) with IPv6 on FreeBSD and
Windows. For IPv4, Windows apparently accepts 239.255.0.0/16 but not
other addresses, so remove IPv4 too.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb682c2839e95d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
QHeaderViewPrivate reimplemented _q_layoutChanged() to handle changes
of rows/columns via layoutChanged/layoutAboutToBeChanged. This worked
fine for Qt4 but since Qt5 only the special signals rowsAboutToBeMoved/
rowsMoved are used for this (8021e2d5e7).
With this change, QAbstractItemViewPrivate::_q_rows/columnsMoved() is
calling the virtual function _q_layoutChanged(). This resulted in a
wrong call of QHP::_q_layoutChanged() for a horizontal header when
a row changed and for a vertical header during a column change. In the
end this can lead to an unhide of hidden sections.
Task-number: QTBUG-54610
Change-Id: Ide4bfc5b24a97746fd1e5af82d3ba08257149157
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
The QTextFormat::FontLetterSpacingType property was added outside
the span of the FirstFontProperty and LastFontProperty, so
the fontDirty flag would not be set when it was changed. There is
no binary compatible way to fix this before Qt 6, so for now, we
add a special case for it.
[ChangeLog][QtGui][Text] Fixed an issue where changing the letter
spacing type of a QTextCharFormat would not cause its font to
update.
Task-number: QTBUG-65345
Change-Id: I5ab53d7f82d529b57edceacfc3fa688c6741cd17
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: C. Boemann <cbo@boemann.dk>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
On Windows, QFileInfo.isWritable() was returning true in situations
where the file would only be writable with elevated privileges. Using
AccessCheck instead of GetEffectiveRightsFromAcl to get the correct
results.
Done-with: Friedemann Kleint <Friedemann.Kleint@qt.io>
Done-with: Edward Welbourne <edward.welbourne@qt.io>
Task-number: QTBUG-30148
Change-Id: I7a3468ac069bf782ca312078e3a84107b6cd468c
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
old style connects have protection against null sender and null receiver,
but new style only had against null sender.
Change-Id: Ie555ac078412918e60c3b60830fe1f3abfb7f5af
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
QHeaderViewPrivate::_q_layoutChanged() was not called when used in a
QTreeView because it was explicitly disconnected in setModel().
The disconnect was added sometime prio to Qt 4.3, but there the signal
was connected to the doItemsLayout() slot. This was correct since
QTreeView::doItemsLayout() is calling header->doItemsLayout().
In Qt 4.3.0 _q_layoutChanged() was introduced and the disconnect was
adjusted. But since _q_layoutChanged() is doing much more than
doItemsLayout() (e.g. restoring hidden sections), functionality was
lost. The problem was already observed for Qt 4.6 (QTBUG-18196) but
only partially fixed.
Task-number: QTBUG-41124
Task-number: QTBUG-54610
Change-Id: Id13a9930d0163812e12a0287016bab9c3aa02068
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Follow-up to d044b0b, with claimed to remove it but
didn’t.
Change-Id: Ifeeb6b062c7c41f1914d677e8aa20e4cb9378ac9
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
If a signal is emitted more than once in a multithreaded application the
QSignalEventGenerator::execute function asserts in the check for a valid
signal index. It happens after abandoning the state and all the
connections are disconnected. If we have pending signal to be
processed the QObject::sender() won't be able to resolve the sender
object.
Task-number: QTBUG-61463
Change-Id: I9d4b7266c6dddc9ff2e7453b05a6989876ccb332
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Commit 68bcccac took account of MS TZ APIs mis-describing non-DST
zones making changes to standard offset (e.g. Europe/Samara in 2011 at
the end of October); however a DST transition that coincides with an
equal and opposite change to standard offset, while the other end of
its DST period changes normally, ends up looking exactly the same
(although it's nominally subtly different), in MS's APIs. Thus fixing
the more common case broke this more obscure case; there is no way to
fix this (other than not using MS's broken-by-design APIs). So kludge
the test to skip the bit we know this breaks.
Task-number: QTBUG-64985
Change-Id: I068500e2e783ab72b400bfd6dbb7dbbd5b08a7bc
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The test would fail on Wayland with window decorations enabled because there
window content is drawn to an FBO before it's blended with the window
decorations (also drawn by the client).
Task-number: QTBUG-51741
Change-Id: I7c74ad235c6a2d50231a9f863da1810675dd4eae
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
Autotest is taken from the previously reverted
8b1377fde1.
Task-number: QTBUG-60231
Change-Id: I44dd79cba22b6baefdd6d95c176790bef0b7eafe
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
QFontMetrics(F)::width() has been deprecated and is replaced by
horizontalAdvance(). This updates all usage of it in tests and
documentation.
It is worth noting that many or most of the usages of
QFontMetrics::width() probably intended to use boundingRect().width(),
but since it currently works, I have not looked into that, just
replaced the function name mechanically.
Change-Id: Iec382e5bad0b50f37a6cfff841bfb46ed4d4555f
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Use being()/end() instead.
These were the last remaining uses of QRegion::rects() within qtbase.
Change-Id: I264beb6f660968f40eecbbee2260341fca94ddb5
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The introduction of begin()/end() actually made most tests
that still use rects() duplicates. Keep them, though, for
test coverage (rects() takes a very different code path
from begin()/end()), but, under the assumption that Qt at
some point will default to built with deprecated APIs dis-
abled, make the test compile with disabled deprecated API.
Port some checks that still used rects() to begin()/end().
Change-Id: Ia7da80f0cd6ef69693f54a474924683624a2c3eb
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Drop the useless copy 'r' of 'region', reuse 'r' for the rects in 'region',
to clean up the region handling in verifyColor().
Change-Id: I8e159976b4b0152823be864488d2ef25d7cb0078
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Test fails to start on FreeBSD:
FAIL! : tst_QUdpSocket::initTestCase() 'networkSession->waitForOpened(30000)' returned FALSE. ()
Loc: [/usr/home/tjmaciei/src/qt/qt5/qtbase/tests/auto/network/socket/qudpsocket/tst_qudpsocket.cpp(234)]
This commit is basically a revert of the Qt 4.8 commit
a951fb79139498774d021759d0466b4b2ff50e68. FORCE_SESSION was only used by
manual testing, as the commit message said
> 8. For manual testing, added the FORCE_SESSION macro to test behaviour
> of UDP sockets when they have an explicit network session associated
So I doubt it has been tested recently.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb569ebf53497b
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
I broke it in commit 4da2dda2aa. It wasn't
flaky or anything: it was plain broken and would never pass. That
indicates no node in the CI has an IPv4 link-local address (and
apparently neither did I at the time).
Change-Id: Ifb5969bf206e4cd7b14efffd14fb62176546916e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
FreeBSD, for example, does not have SO_NREAD and its FIONREAD returns
the full socket buffer size, including IP and UDP headers.
Change-Id: Ifb5969bf206e4cd7b14efffd14fb5d8ca778d16a
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Also drops a few instances where the dependency was purely runtime,
especially for examples.
Change-Id: I2a0476f79928143596bdb3b8f01193af90574ae8
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
As the numercialPrecisionPolicy can be set and subsequently retrieved
from the QSqlDatabase's driver, then when copying the QSqlDatabase, we
need to set that appropriately too.
Task-number: QTBUG-10452
Change-Id: I2c63748365ab4e9fbc29d8d460d80d2e2a0ee385
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
We haven't had support for anything older than 7.3 for a long time, so
this removes the code to save having to maintain it any further.
Task-number: QTBUG-644
Change-Id: If0635f3bef5138a00a0e77011a70f23d0bffda32
Reviewed-by: Robert Szefner <robertsz27@interia.pl>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
A mysterious file c:\dev\null appeared when running the auto tests on
Windows. Fix by using QMAKE_SYSTEM_NULL_DEVICE.
Amends 5f3529be32.
Change-Id: I0224a9ccd61c4f10b2ddb8f8d690e1849aa88d8a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The test is flaky on Linux.
Task-number: QTBUG-64639
Change-Id: Iec56ebce4f656f52187b34c8f655b137e41c3d17
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
QAbstractItemDelegatePrivate::textForRole() was formatting date and
time separately, then gluing the parts back together with space.
QLocale can do that just fine itself (it has a toString() overload
that handles a QDateTime) and might even (some day) do it better. To
my mild surprise, this proved sufficient to fix a problem with
date-time display in tool-tips, when the date-time includes a zone.
Extracted the date-time part of an existing selftest into a test of
its own and extended it to test times of each spec-type; verified that
the non-local spec cases of this all failed before this fix.
Task-number: QTBUG-61069
Change-Id: I6d6be0c27be9a557d8afc3ced200a10b2aaff816
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Previously, a 1, 2 or 3 for "dd" would be rejected because 10, 20 or
30 would fit in the field and be valid; but 4 or more was accepted,
even though it was too short for the field, because no suffix could
make it valid within the field-width.
[ChangeLog][QtCore][QDateTime] When parsing dates and times from
strings, fixed-width date-time fields, such as a "dd" for day,
QDateTime now rejects all values that should be padded, rather
than only doing so when the value is a prefix of some value that
would fill the field-width. Use a single letter for the field,
e.g. "d" for day, if you want to accept short
values. (QDateTimeEdit is not affected.)
Task-number: QTBUG-63072
Change-Id: I22d223c50057c3edab4ef7f01d9ed0f58e9139c1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Test-case taken from bug-report; fits in as an easy row in an existing
data-driven test. Add similar tests for date-time and time; and an
isValid test on the end of year 9999. The date-time parser was using
the end of year 7999 as maximum value for dates and date-times; extend
this to year 9999, as I can see no reason not to.
[ChangeLog][QtCore][QDateTime] Years up to 9999 can now be parsed
without error (previously 8000 and beyond were treated as invalid) in
all formats (not only in ISO format). Widgets handling dates now
support dates to 9999, likewise.
Task-number: QTBUG-64401
Change-Id: I518cfa6c2cb4ecc5a85b896dc9e56b4fdd8a8bb1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When a new source model was set to QSortFilterProxyModel, the model
tried to remap the persistent indexes to the new model which was wrong.
The correct solution is to clear the persistent indexes with
_q_sourceModelDestroyed() since the old source model went away.
Task-number: QTBUG-44962
Change-Id: Id39e9ac83324250e8bfa434aae467a9206d2590e
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
setSession is only used if we have bearer management.
Change-Id: I64b9d29c01566e79bbca5d0dc11d6aee6d9b0bf0
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
QTreeWidgetItem::insertChildren should behave like a more-optimized
QTreeWidgetItem::insertChild. Unlike the latter, the former lacks
out-of-bounds checks, resulting in successful insertions even
when using an invalid index (say, bigger than the row/column count).
Reintroduce some sanity checks instead. This allows to fix a "fixme"
left in the autotest.
[ChangeLog][QtWidgets][QTreeWidgetItem] QTreeWidgetItem::insertChildren
now ignores insertions happening at invalid indices, for
consistency with QTreeWidgetItem::insertChild.
Change-Id: I1532597768cc6aff96a6e8f356bc6075b582801d
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
The previous commit didn't handle correctly the case where an entire
mime directory is deleted. The unittest wasn't testing that case,
now it is. We need to move providers into a new list, and then
delete those left over (i.e. now unused).
Change-Id: I04fd8b39b511a2331d706864f695ce5074acf916
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Previously, we would use mime.cache in all mime directories if at
least one of them had such a file (other than the most-local one),
otherwise the "source" XML would be used in all directories.
Now it's possible to use mime.cache in those directories which
have one, and XML in those directories that don't.
Not only is this more correct, it will allow in a subsequent
commit to bundle the binary cache in QtCore's qrc rather than
the very big XML file.
The design change to allow this is that now every provider
only deals with a single directory, and QMimeDatabasePrivate
takes care of creating multiple providers, one for each dir.
This required to move most of the loops from the binary provider
up to QMimeDatabasePrivate itself.
Change-Id: Iac82d5f5836b80d45076e586b903d16fa2525b34
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The tests start to show flakyness on Linux in 5.10 (rowCount() check
failing). This seems to point to a race condition between the files
showing up and the file system watchers of QFileSystemModel starting.
To fix this, close the file and wait until it shows up in the directory
before pointing the QFileSystemModel to it. The tests then no longer
rely on the file system watchers.
Change-Id: I39cffb4cacf6843e8e4180efb405345307c78dd8
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@qt.io>
The reason for each is given in the skip. It's mostly about the
server-side encryption, which is unimplemented for WinRT.
Change-Id: I036b95a4526e02fd047e193f2b3c9130bec08144
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
This lets the tests run on devices which previously did not have access
to the files used (WinRT, mobile devices).
Change-Id: Ibdd85862eee6ab1a7d4da87ca321ee9bc9880bfa
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Since removeAll() takes its argument by cref, if passing a reference
to an element of the container to removeAll(), the element may be
deleted (overwritten) by anyother value, leading to UB.
Add a test that actually happens to fail for me without the patch,
even though that might not be guaranteed (we may invoke UB).
Change-Id: If8c795113aeb515f4a9bdf1e072395b932295667
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
During the container BoF session at the Qt Contributor Summit 2017 the
name of the signed size type became a subject of discussion in the
context of readability of code using this type and the intention of
using it for all length, size and count properties throughout the entire
framework in future versions of Qt.
This change proposes qsizetype as new name for qssize_t to emphasize the
readability of code over POSIX compatibility, the former being
potentially more relevant than the latter to the majority of users of
Qt.
Change-Id: Idb99cb4a8782703c054fa463a9e5af23a918e7f3
Reviewed-by: Samuel Gaist <samuel.gaist@edeltech.ch>
Reviewed-by: David Faure <david.faure@kdab.com>
We have had test.qt-project.org for close to 3 years now.
Change-Id: I71488efd29b645f7b228fffd14fadf4627288243
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@qt.io>
(cherry picked from commit 5f66f87181)
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
We have had test.qt-project.org for close to 3 years now.
Change-Id: I71488efd29b645f7b228fffd14fadf4627288243
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@qt.io>
Allows categorized logging before QCoreApplication has been created,
which otherwise would silently fail to output anything because the
category would never be enabled, despite QT_LOGGING_RULES being set.
Change-Id: I1861e5366ea980dff2ffa753b137276c77278eee
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
The test is removing 2 rows starting at the last row of a model.
As the comment indicates, that's clearly invalid.
Change-Id: I43ef00d602934965b206e2ba591ff8fd0a6ae398
Reviewed-by: David Faure <david.faure@kdab.com>
Unary ~ is not defined for enum classes, so we need a cast.
Change-Id: I79d495ebcc24ab960da8dae3be08eb307a9de448
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... and unblacklist it on Windows.
From what I can tell there is no particular reason why this test fails
other than that it is a little too slow sometimes (these things happen).
So, to fix the test I bumped the timeout, but to avoid the test running
for longer on every test-run it now also ends when the socket enters
the "Unconnected" state.
Previously it failed 171/500 times, and after this patch it failed
0/1000 times.
Change-Id: I4266bff6b91aaaf502ee66265d01c3a177706402
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
The assertion in isUnnamedFile() we had was incorrect after the file was
removed, since we cleared the name and possibly reset back to the
template. Since ~QTemporaryFile() calls remove(), this was easy to
trigger if you attempted to remove the temp file and leave
QTemporaryFile like that.
Take this opportunity to add to the docs of setAutoRemove() explaining
the possibility of unnamed files.
#7 0x00007f69bcc2b50e in qt_assert (
assertion=assertion@entry=0x7f69bcf194a0 "unnamedFile == d_func()->fileEntry.isEmpty()",
file=file@entry=0x7f69bcf19458 "io/qtemporaryfile.cpp",
line=line@entry=514) at global/qglobal.cpp:3123
#8 0x00007f69bcd672cf in QTemporaryFileEngine::isUnnamedFile (this=this@entry=0x55cd60644df0)
at io/qtemporaryfile.cpp:514
#9 0x00007f69bcd683f7 in QTemporaryFileEngine::remove (this=0x55cd60644df0)
at io/qtemporaryfile.cpp:396
#10 0x00007f69bcd48654 in QFile::remove (this=this@entry=0x7fffb393f7e0)
at io/qfile.cpp:513
#11 0x00007f69bcd6653b in QTemporaryFile::~QTemporaryFile (this=0x7fffb393f7e0, __in_chrg=<optimized out>)
at io/qtemporaryfile.cpp:719
Change-Id: I57a1bd6e0c194530b732fffd14f4ed28ca8185b2
Reviewed-by: Andreas Hartmetz <ahartmetz@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
A layoutChange indicates that anything can have moved to anywhere else,
including as a result purely of new items being added. It can also
indicate that items are removed.
The old code here incorrectly assumed that the section count remained
constant over this operation by setting the size of the oldSectionHidden
QBitArray - whose size is the size before the layoutChange operation -
and then calling setBit with model rows numbered after the layoutChange
operation. As the two are not necessarily the same dimensions, this can
result in asserts from the setBit call.
Simplify the handling of layoutChanged entirely by clearing section
information, and using the QPersistentIndexes which indicate hidden
state to restore that state after re-population.
Task-number: QTBUG-53221
Change-Id: I3cda13e86b51b3029b37b647a48748fb604db252
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Make sure we don't run into warnings for CMake 3.10
Task-number: QTBUG-63442
Change-Id: Ida004705646f0c32fb4bf6006036d80b1f279fd7
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Sebastian Holtermann <sebholt@xwmw.org>
Reviewed-by: Rolf Eike Beer <eb@emlix.com>
... and unblacklist it.
It was blacklisted some years ago because it was failing too often.
It was failing because the ssl socket had already received and decrypted
all the data it was going to get, meaning the waitForReadyRead call was
just going to block forever.
Change-Id: Ia540735177d4e1be8696f2d752f1d7813faecfe5
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Since the first item in a treeview might be hidden, start from the
first visible item in the view when starting or wrapping round
during a keyboard search.
Task-number: QTBUG-63869
Change-Id: I202bea567c6d4484c3ffaf8a5f9af8ea2e13708d
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
This test became a real pain recently. A close look at the test shows
several problems (strangely enough, the failure can never be
reproduced on real machines, only on VM - Ubuntu and RHEL 6.6).
There are several asserts that are firing from time to time here and
there. They show that the logic in test is broken/incorrect. QNAM can
open several connections to a host, our test then incorrectly resets
its 'client' data-member and bad things can later happen after
'bytesWrittenSlot' executed (and deleted a socket). For example,
I can reproduce this scenario in every second run:
1. incoming connection -> client = socket(descriptor), connect to
client's readyRead (s1)
2. incoming connection -> client = socket(descriptor), connect to
client's readyRead (s2)
QNAM sends a request on s1. We reply on s2 (which is already wrong)
and call client->deleteLater(), which resets client to nullptr.
If QNAM sends something else on s1, we hit assert(!client.isNull()).
To avoid this, whenever 'sender' in any slot is different from the
'client', we use the actual 'sender' to reply. Another problem is this
weird and rather cryptic waitForFinish which is not needed in this
particular test since we wait for reply error, not 'finished'.
As it happened before - it's not clear if these two problems
were the cause of guaranteed fails on CI - an integration failed
~10 times in a row in the same test (not happening anymore though).
Task-number: QTBUG-64569
Change-Id: Id9aa091290350c61fadf1c3c001e7c2e1b5ac8f4
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Both ARM and x86 can convert fp16 much faster in bulk than one at a
time. This also enables hardware accelerated conversion on x86, when
F16C isn't unconditionally available at compile time.
This code is implemented in C to ensure that there's no leakage of
inline symbols from the .obj file that was compiled by Visual Studio
with AVX support. Unfortunately, simd.prf uses $(CXX) instead of $(CC)
for all its sources, which means the file gets interpreted as C++ by
g++, clang++ and icpc. Those compilers at least don't leak any symbols.
Done-with: Thiago Macieira <thiago.macieira@intel.com>
Change-Id: I9d26d99e83392861fb09564e0e8e8d76cd8483b3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Tested with Clang, GCC 4.5 & up, ICC 17 and MSVC 2017. No current
version of MSVC supports C11 and GCC implemented the features slightly
later in C than in C++.
Change-Id: I57a1bd6e0c194530b732fffd14f45c5074c9a052
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
When calling QTreeView::hideColumn() the row heights are not
recalculated. This can lead to rows which are unnecessarily high due to
hidden columns may contain large (e.g. multiline) content. The same
applies to showColumn() - there the row might be to small.
Hiding columns directly via QTreeView::header()->hideSection() is not
covered by this patch since QHeaderView has no way to inform about
newly shown/hidden sections.
Task-number: QTBUG-8376
Change-Id: I20b1198e56e403ab8cf649af76e5e2280821dd68
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Also blacklist tst_QNetworkReply::ioHttpRedirectErrors(too-many-redirects)
on RHEL 6.6 in CI.
Conflicts:
tests/auto/network/access/qnetworkreply/BLACKLIST
tests/auto/network/ssl/qsslsocket/tst_qsslsocket.cpp
Task-number: QTBUG-64569
Change-Id: I7514fc0660c18fd3a3e1d0d0af3f15d879e3c6f4
This test makes sure that we do not introduce a regression where the
threads exited the inner loop over the queue before the queue was
empty. This was triggered by calling tryTake at least maxThreadCount
times, which left the same number of null pointers in the queue
and caused the inner loop to exit too soon for all the threads.
Change-Id: I3a9d800149b88d09510ddc424667670b60f06a33
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The verifySessionProtocol() method in the SecureTransport backend did not
properly handle TlsV1_0OrLater, TlsV1_1OrLater and TlsV1_2OrLater.
This commit teaches verifySessionProtocol() about them.
It also adds TlsV1_0OrLater, TlsV1_1OrLater and TlsV1_2OrLater to the
protocolServerSide() test in tst_qsslsocket.
Backport from 5.10 to 5.9 (LTS).
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
(cherry picked from commit 9c765522d1)
Change-Id: I58c53bdf43e0f19b4506f3696d793f657eb4dc6f
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
As it stood, we would set 'pressed' to false regardless of which button that
was released. This would end up wrong if pressing the left button, and
at the same time, did a click with the right button. This would clear the
flag prematurely, and cause a release signal not to be emitted when later
releasing the left button.
tst_QAbstractButton: adding autotest
Adding tests to simulate the bug report's cases:
1) left press button
2) click right/middle key
3) move mouse out of button's boundary
4) test if the released() signal triggered properly
Taks-number: QTBUG-53244
Change-Id: Ifc0d5f52a917ac9cd2df5e86c0475abcda47e425
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
