Change-Id: If844785d7d3645c33e0fcb1206cc52f8ab644070
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Alexander Neundorf <neundorf@kde.org>
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Previously, ignored messages were only cleared at the end of each test
function, i.e. after all data rows were finished. This meant that if a
data row in a data-driven test function didn't cause all of the expected
messages to be generated, the remaining messages would be carried over
to the next data row. This would result in errors about missing
messages being associated with the last data row rather than with the
correct data row.
This commit makes testlib check for missing ignored messages after
running each data row rather than only doing so after the last data
row.
This commit also adds a regression test to demonstrate that ignored
messages can no longer be carried over from one data row to another.
Change-Id: Ibee51aa6e96866fbcbcb4acee1a8340a86a6a4ba
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Previously, expected failures were only cleared at the end of each test
function, i.e. after all data rows were finished. This meant that if a
data-driven test function called QEXPECT_FAIL and then didn't perform
any further verification steps to trigger the expected failure, the
expected failure would be carried over to the next data row, probably
causing the first verification step in the test function to XPASS (with
a seemingly irrelevant error message) for the next data row.
This commit adds the new function QTestResult::finishedCurrentTestData()
to cleanup after each data row is executed. This function treats calls
to QEXPECT_FAIL without subsequent verification steps as a test failure.
This commit also adds a regression test to demonstrate that expected
failures can no longer be carried over from one data row to another. If
run against the previous version of testlib, the new test would report a
pass instead of an error.
Change-Id: Ida5c7f080815b0dca9531131fed582b0918334cb
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
linuxinput becomes evdevmouse. The experimental touch code is removed,
now the plugin's purpose is solely to generate mouse events from
absolute and relative pointer events. The plugin key is EvdevMouse.
touchscreen becomes evdevtouch. The plugin key is EvdevTouch.
In case keyboard support appears some day, it will fit nicely in the
system by the name of evdevkeyboard or similar.
Some little udev code is moved to platformsupport so it can be shared
between the plugins. This may be extended later if more sophisticated
udev support is needed. N.B. the intention is to keep this as simple
as possible. We are shipping these plug-ins as reference examples, not
as full-featured drivers.
evdev and udev support has configure time tests from now on. This
means the "drivers" (generic plugins) will get built automatically
when the support is available.
Change-Id: Iaf6260b5c2edfb9f25d070d2764466725adc6b4e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
QWidget::isActiveWindow() was being checked in QAccessibleWidget::state
to determine if a widget is focusable. As a result, focusable widgets
were reported to be not focusable when the window was not active.
Change-Id: I73c47181ed132a84f0251cb67d0e20912e29a1a6
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@nokia.com>
Timers are not entirely precise; if we ask for a timeout of
10000 milliseconds, we might time out in 9999 instead.
Also, we know the expected elapsed time in each case, so do a fuzzy
comparison against that time.
Previously the test was verifying that the elapsed time was greater than
or equal to the timeout in the case where a timeout was expected, which
means the test would not detect bugs which incorrectly caused the
timeout to occur later than it should.
(cherry picked from qt4 commit 9a2573dc13b3e8df6cd15bef64370ea407480fc7)
Change-Id: I91d0c81f989ab43a3c48f6abbb4c5b28e2b35402
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
This autotest assumed that various network operations could always be
completed within 5 seconds. Notably, it assumed that an attempt to
resolve a nonexistent hostname would always result in an error within
5 seconds. In my testing, it usually takes 4-6 seconds to complete,
but occasionally takes as much as 13 seconds.
(cherry picked from qt4 commit c85faef67b6a7e8fcedb4ce800282d41f5b79ec1)
Change-Id: I982ecf6ebc1bb8ee2184cf5592cb2684474c870b
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
If an invalid range of rows is specified, it's likely to be a
programming or user error. The old behavior of ignoring out of range
rows seems dangerous and complicates the code.
Also implement the documented behavior of returning false if
changes are unsuccessful for OnFieldChange and OnRowChange.
Previously the return value of submit() was ignored.
Updated and improved documentation.
Change-Id: Iaaf51c6d9a0c8c06fd5d186b4b88358fbeab9936
Reviewed-by: Yunqiao Yin <charles.yin@nokia.com>
Qt 5 seems like a welcome opportunity to stop emitting this
spurious beforeDelete signal.
Change-Id: Ib8628343ca9b8fdd85c154a206c7e2bf2c4c9dc1
Reviewed-by: Yunqiao Yin <charles.yin@nokia.com>
Even though ModifiedRow is not part of the public API, guarding its
data helps make clear the intended patterns of use.
"op" and "primaryValues" are read-only after construction.
setValue() encourages maintainers to let ModifiedRow manage the
"generated" flags of the record.
The primeInsert() signal still exposes the actual record, including
the "generated" flags, which is the justification for recRef().
Change-Id: I16d1610a8f9233af78b90662b08706b48ea19c41
Reviewed-by: Yunqiao Yin <charles.yin@nokia.com>
First, heed the warning from qmake and add load(qt_module) to the
top of src/concurrent/concurrent.pro
Second, qtconcurrentversion.h is in src/concurrent, not src/xml
Change-Id: If4fc55ac2fe21c093023b377195ed82a54d8c963
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Window system defines have been deprecated,
so use Q_OS_WINCE instead.
Change-Id: I52059d0f854fe783ac20610ab248800c3e1e827c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Fix compilation with MinGW-w64: Use quintptr instead of
unsigned long as WId needs to be able to cast to
HWND which is a 64-bit pointer on Windows 64-bit.
Acked-by: Samuel Rødal <samuel.rodal@nokia.com>
Change-Id: I7bbe83ce02b739eb04218ca4b6b478509347d515
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The file was moved, but got never renamed.
Change-Id: I4f13d87976cf35a1e14b9a8c178050332f45b7ee
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@digia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The open test case tests opening a device path on windows.
As this requires privilege elevation, the test fails when running
nmake check from a console.
As the CI machines appear to run tests with admin privileges, first
check opening the device using the windows API. If that succeeds, then
opening using QFile should succeed. If that fails, the opening using
QFile should fail.
(Since Windows vista, members of the "administrators" group do not run
at elevated privilege all the time. The user needs to opt in via a
UAC prompt or "run as administrator". This is similar to using the
sudo command on unix)
Ran the test as administrator and normally. Now passes in both cases.
Change-Id: Ibd7682eceb61e35d4912fa0392df536f4331f6ed
Reviewed-by: João Abecasis <joao.abecasis@nokia.com>
QtUITools is a static library, so we need to mark that in its
CMakeConfig.cmake file.
Change-Id: I4e3684f7c62f4af28956a6e8f58edce7724f50dd
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Simplified the raster engine by treating image formats in a more
generic way and removed some unused code.
Change-Id: Ib3979a1a6e3e6f17c5002248545779ec36fff7c9
Reviewed-by: Gunnar Sletta <gunnar.sletta@nokia.com>
There is no mktemp for Windows CE, nor mkdtemp. So use the
implementation for QNX for Windows CE too, to simplify
the implementation use it for windows too.
Change-Id: Icddaf474a2d696752d8f9774fb4f033454e34e83
Reviewed-by: David Faure <faure@kde.org>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
this commit is aimed to fix QTBUG-4341. now QFactoryLoaderPrivate's
destructor will call unload() to QLibaryPrivate object which will destory
the plugin's root instance if its refcount reach zero.
Task-number: QTBUG-4341
Change-Id: I3cd3e071b34271bf5802ab09f6c125beda5e9844
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
Reviewed-by: David Faure <faure@kde.org>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Commit b7ca6a81db removed a call to
sendPostedEvents() that deemed unnecessary. Unfortunately, it is
necessary, as shown by the
tst_QScriptEngine::processEventsWhileRunning() test in the QtScript
module. Re-add the call, but only when not waiting for more events.
Change-Id: I648d66dd3ba484ad9e9a93fc03a9792cca5035c6
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@nokia.com>
Currently when tests are crashing, aborting or hanging, their status are
reported with a status "Unknown" because SaveCoverageTool is never called
for them. The status of the test given by the coverage tool should be the
same as the one reported in the log output when the test is run. This change
will allow more accuracy in the status reported.
Add QTestResult::setCurrentAppname and QTestResult::currentAppname.
To retrieve the name of the current application running.
Task-number: QTQAINFRA-460
Change-Id: Icc476dc2d6cb28185e5447f1e79da6a8a31cad54
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Reviewed-by: Jason McDonald <jason.mcdonald@nokia.com>
QMetaType used to register a typeName and factory functions for
creation/destruction of objects. While it would be possible for a single
type name to be registered matching different actual types and memory
layouts, there was little that could be done about it.
Now that QMetaType is tracking type information with a direct impact on
data layout and ABI (size and type flags) it is important that we check
and detect binary incompatibilities as early as possible.
[Such incompatibilities could arise from type name re-use (technically,
ODR violations) or, more commonly, as version mismatch between different
shared libraries or plugins.]
Only type size and flags are checked as function pointers to inline and
template or otherwise non-exported functions could trivially differ
across translation units and shared libraries.
When registering typedef types, a check is made to ensure the same name
doesn't get registered as different types.
Change-Id: I8211c3de75d4854ce8fafdb620d3a931c206e0c3
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@nokia.com>
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
The function hasn't been working properly. It was not well tested, for
example it is undefined how QVariant should behave if it contains an
instance of an unregistered type.
Concept of unregistering types was inspired by plug-in system, but in
most supported platforms we do not unload plug-ins.
Idea of type unregistering may block optimizations in meta object
system, because it would be not possible to cache a type id.
QMetaType::type() could return different ids for the same name.
Currently QMetaType::unregisterType() is not used in Qt.
Change-Id: I878b6e8d91de99f9bcefeab73af2e2ba0bd0cba0
Reviewed-by: Prasanth Ullattil <prasanth.ullattil@nokia.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: João Abecasis <joao.abecasis@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
There is no point in keeping separate values which should mean the
same.
QVariant::UserType was used also to construct a valid, null QVariant,
containing an instance of unknown custom type. The concept was strange
and useless as there was no operation that could be done on such
QVariant. Therefore it was dropped.
Please note that the patch slightly changes behavior of different
functions accepting a type id as parameter. Before QVariant::UserType
was an invalid type from QMetaType perspective (id 127 was not assigned
to any built-in type), but QMetaType::User points to the first registered
custom type.
Change-Id: I5c7d541a9affdcdacf53a4eda2272bdafaa87b71
Reviewed-by: Kent Hansen <kent.hansen@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: João Abecasis <joao.abecasis@nokia.com>
Reviewed-by: Andrew Stanley-Jones <andrew.stanley-jones@nokia.com>
Reviewed-by: Aaron Kennedy <aaron.kennedy@nokia.com>
The QNetworkReply finished signal does not have a bool parameter.
Change-Id: I87bd0410545f7a2fc2ab63cca90548f0585bf7a0
Reviewed-by: Shane Kearns <ext-shane.2.kearns@nokia.com>
Reviewed-by: Jonas Gastal <jgastal@profusion.mobi>
QTest::ignoreMessage() cannot be called when creating test data rows,
but rather must be called when executing a row.
Change-Id: Ic958c4e5c15dd53a48479099b6b07803af6cb789
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
QTest::ignoreMessage() cannot be called when creating test data rows,
but rather must be called when executing a row.
Change-Id: Ifceaff4c5a7f1b6408ec57196298e3f3038910c9
Reviewed-by: Rohan McGovern <rohan.mcgovern@nokia.com>
Remove the #if 0'd blocks and instead add these headers to
sync.profile's @ignore_for_include_check, as documented at
http://wiki.qt-project.org/Creating_a_new_module_or_tool_for_Qt#other_fields
Change-Id: I3bd6e8cb21eca139fdca10fe970eeaf2e4e77c24
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
When closing a window, destroy() ensures a setVisible(false) call only
for the window itself, not for other windows parented to it. With the
new quit lock ref feature this breaks code that creates a fake root
window parented to the main, visible window. (for example the
xcomposite backends of qtwayland do this)
Such apps do not anymore exit after closing their window because the
children do not receive a deref due to setVisible not getting called.
(At that point. It would get called after exiting the event loop but
that never happens due to the refcounting)
Change-Id: I124737c80ad59600ddc79261100f3904af0f410d
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
The code was hard-wired to use QFSFileEngine to determine platform
preference for file system case sensitivity. In this case it is cheaper
to use private API and directly access the information in
QFileSystemEngine.
This change is also necessary because file engines are being dropped
from the public API.
Change-Id: I5015d5fdb3979af9ff2d114084053ad06220d834
Reviewed-by: Shane Kearns <ext-shane.2.kearns@nokia.com>
As far as Qt is concerned, this information is static and depends only
on the platform. By moving the definition to the header we allow the
information to be directly used by the compiler, forgoing the need to
export the function to interested users.
Change-Id: I43de585391d41204d3c3560ac6e65e5a88aee11c
Reviewed-by: Jonas Gastal <jgastal@profusion.mobi>
Reviewed-by: Shane Kearns <ext-shane.2.kearns@nokia.com>
Delete QFontEngineBox object in QFontDatabase::load() if the object can not
be used.
Change-Id: Ifb055809703fb6be92f41fd56658198df30837fd
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@nokia.com>
Since we may use distance field text rendering in raster, these
functions need to be moved from declarative to QtGui.
Change-Id: I158bc3bae02b5590ae812f06ad7a78a5914bcb9e
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@nokia.com>
Most of these headers are either forwarding headers, or we explicitly
stop syncqt so that it doesn't generate class includes for the atomic
implementation. Either way, syncqt doesn't see the QT_END_* (and
sometimes not QT_BEGIN_*), which this commit fixes.
Change-Id: Icc8da6f384f38b1ff4eb265c731ce2f2ed92a1a3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This silences the warning from syncqt:
QtCore: WARNING: qtbase/src/corelib/global/qglobal.h includes qlogging.h
when it should include QtCore/qlogging.h
Change-Id: I64bd92898190031eb0d3d1dfa5ba5bc56db01c00
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qcompilerdetection.h is included from qglobal.h, and does all Q_CC_*
and compiler feature detection.
Change-Id: Idd06054e172ef6fa73774e26fa38753996c4161b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qsystemdetection.h is included from qglobal.h, and does all Q_OS_*
detection
A side-effect of this change is that QT_BEGIN_HEADER and QT_END_HEADER
started being defined on Mac OS X, which ends up breaking the build
in some cases. Since QT_BEGIN_HEADER and QT_END_HEADER have been
defined to nothing in the past, even on Mac OS X, change these 2
to be unconditionally defined to nothing.
Change-Id: Ibc8a0aa2207664741c25627d7621e006c2ce80d3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Exclude invisible top level dialogs for whom a native
dialog is being shown from top-level search.
Change-Id: Ia94599905457d81d342c14d09ad0b0fc89ec4ab1
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
