e.g. for the text "Aaa bbb ccc.\r\nDdd eee fff." the \r\n wasn't treated
as a hard line break or as a line break opportunity, ever.
Quite ancient bug...
Change-Id: I8d8497c55a3a4d51c27de99ccfe1e31f3bf4de77
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Add QUnicodeTools namespace and rename qGetCharAttributes to initCharAttributes;
Make it possible to disable tailoring globally by overriding
qt_initcharattributes_default_algorithm_only value
(useful for i.e. running the specification conformance tests);
This is mostly a preparation step for the upcoming patches.
Change-Id: I783879fd17b63b52d7983e25dad5b820f0515e7f
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Some names were fixed to meet their values in the Unicode data files (e.g. Zain -> Zhain);
Some other names were added as a value alias (Nya -> Noon) or deprecated
but retained as a value alias for BC (HamzaOnHehGoal -> TehMarbutaGoal);
Newly assigned Arabic characters were mapped into respective joining types and classes)
Change-Id: Ic4b4a6baf70f51fb5936f74fe9a173b8790dda11
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
It's only used internally in qwidget, so it might as well
live in qwidget_p.h.
Task-number: QTBUG-25070, QTBUG-25373
Change-Id: I87770e0b49253b4bdebe90ce84dd42448d6175bd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This adds some pure virtual functions to the interface to enable
batching requests.
The interface functions no longer need to be invokable, as
invokeMethod() is too slow to be used at the required frequency
anyway.
Change-Id: I89b607de1a28f3f4b728c334c04bcd443d4366b4
Reviewed-by: Jiang Jiang <jiang.jiang@nokia.com>
* Merge the two public ctors.
* Use bitflags instead of shifting bits (more readable).
* Add autotest
* Use int datatype for the "stretch setters". (values out of bounds are clamped)
Streaming to QDataStream will still use the Qt 4 format.
Task-number: QTBUG-25100
Change-Id: Iecb1e78cb12717e4d84448484c3ad8ca469d571a
Reviewed-by: Paul Olav Tvete <paul.tvete@nokia.com>
Upgrade to new gbm_surface API which lets us use eglCreateWindowSurface.
Change-Id: I4af86d442375017ae3b4744e7b3bdebf3b496b6e
Reviewed-by: Andy Nichols <andy.nichols@nokia.com>
A previous commit changed the Mac behaviour for printerName()
from returning the CUPS Description to returning the CUPS Name.
In case anyone was relying on this for a human-readable name
add new api to return the CUPS Description. Also add the
Location and Make and Model which will be used in the Unix
print dialog instead of directly calling CUPS.
Change-Id: I9901bf8d6368466adf111580f5db5a3f01ca9170
Reviewed-by: Teemu Katajisto <teemu.katajisto@digia.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: John Layt <jlayt@kde.org>
QWindow::setSurfaceType and QWindow::setFormat do not re-create
the QPlatformWindow. Attempt to document this behavior and point
to the documentation of destroy() and create().
Change-Id: Idf7eb343d4918a45b5a701effe3263145a33790a
Reviewed-by: Casper van Donderen <casper.vandonderen@nokia.com>
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Change-Id: I7df089a9dc87396e19c08cef3d76c7c7f95a8fbf
Reviewed-by: Donald Carr <donald.carr@nokia.com>
Reviewed-by: Andy Nichols <andy.nichols@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
The eglfs backing store code can render either using the raster paint
engine (through QImage) or using the opengl paint engine. Rendering
quality using the opengl paint engine is pretty poor so remove
it.
Change-Id: I64061ceb3a480049cfebe61aaf172ad1f1da7042
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Change-Id: Ifa096537d417571ed660a63a501fed188186f400
Reviewed-by: Donald Carr <donald.carr@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
\#include "foo" in MSVC does not search the output path, so we need to
pass the proper -I flag to find the build dir's src/corelib/global.
Change-Id: I546051c99fd29d7734f8bf35d058ea283ec9bca7
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
QWidgetPrivate::drawWidget was missing an important line about marking
native widgets dirty compared to 4.8, which caused all other widgets
besides the native one to not repaint unless window was resized in such
a way that the widgets would get hidden and then re-exposed.
The above didn't fix repainting issues when moving native widgets
(e.g. widgets in QMdiArea subwindows or just calling QWidget::move()).
Added setting widgets dirty to QWidgetWindow::handleExposeEvent to
address this issue.
If there is one native widget, Qt enforces that all widgets in same
parent hierarchy are native - presumably to get overlapping widgets
drawing correctly. However, qapplication_qpa.cpp set the attribute
Qt::AA_DontCreateNativeWidgetSiblings for all applications, which
caused only the parents of native windows be forced native, leading to
drawing artifacts related to siblings of native widgets (e.g.
overlapping QMdiArea subwindows). I don't see a reason for setting this
flag indiscriminately for all applications, so removed it.
Also added setting newly created QWindow visible if associated
widget is already visible, which can happen if regular widgets
are shown before the first native widget is shown and retroactively
forces other widgets native.
Task-number:QTBUG-25805
Change-Id: Ib133dae9b13cc6e7155e7cae00fc1339d3b5ae86
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
This fixes a remaining feature regression from Qt 4.x, where the
dragmanager also determined the default action for a drag given
the set of supported actions and the keyboard modifiers.
Task-number: QTBUG-25373
Change-Id: I80b23b135ba218eb5b7ccc692f12140477bc7809
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
The documentation is misleading in that a default-constructed
rectangle that has size = 0 causes the entire widget to be
painted; this happens only if the one of the size values is
< 0.
Add a default value such that size < 0 and the entire widget is
painted based on the rationale that a size = 0 might be the result
of some calculation and it would be unexpected to get a pixmap
containing the entire widget in that case.
Change-Id: I2a5fb38cd407b4b64348213101d8196b2c40ad9a
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
This breaks the combo dropdown animation on Windows as it
uses special layout tricks and potentially other (unlaid-out)
widgets.
If the size is unknown, call prepareToRender instead.
Reverts d6f971c854.
Change-Id: If44e2a7e7fb134c228674cf228bbee0d5520e04f
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Prevent the classes from appearing in the Qt documentation.
Change-Id: I62abff0c57effdf16629f9d5a0dc384ea2c43d5a
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@digia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
libudev and linux/input.h are Linux specific. Instead of add a !*bsd*
to the pro line I have decided to only check for linux-*. This is fixing
the compilation on FreeBSD.
Change-Id: If59481421e811eca205407031db593e00b6e8534
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Try ensuring the window has received its valid position from the WM.
Change-Id: Ibd75bc19ae820765bfaadd30c22e77a19cd28849
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
This is a work-around for a feature missing in Harfbuzz that
can make a text run that contains a surrogate and a ligature
crash. This will potentially cause the ligatures to break up
if you combine them with a surrogate, causing visual changes
to the text, but the scripts that require GSUB should not be
affected by this, since they will not use surrogates. Still,
it's not a permanent fix, but will serve as a bandaid for the
crash until the underlying problem has been fixed.
Task-number: QTBUG-22275
Change-Id: I90c37fba76bc7d1f369f3afddd1bd0dc306f5750
Reviewed-by: Jiang Jiang <jiang.jiang@nokia.com>
qiterator.h doesn't use std::bidirectional_iterator_tag or
std::random_access_iterator_tag, so remove the forward declarations.
(These actually confuse clang when building with C++11 support enabled).
Change-Id: Idd0daa0840a0995e6e5b0ffc01ddcdf06f048149
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Unlike most other modules, QtDBus in its use of QVector also needs to
check if the item isn't null, which makes for some special changes.
Change-Id: Ia22ad2a6b26c9c34dc09ab882d81323a941d166a
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
We'd like to decrease the default timeout for tests in the Qt Project CI
so that we waste less time waiting for hanging tests.
Tests which genuinely take a long time to run, such as these, should
have their timeout explicitly set in their .pro file.
Change-Id: I4fe6249e9efa764b230251d73a1115c24411e168
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Reviewed-by: Kalle Lehtonen <kalle.ju.lehtonen@nokia.com>
Use a QTemporaryDir instead of the system-wide temporary directory.
The test is still not entirely parallel-safe (at least on X11) due to
requiring the shown dialog to have keyboard focus.
Change-Id: I628dc6ab52dda49f6957a301eea8944bb9d81453
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Reviewed-by: Kalle Lehtonen <kalle.ju.lehtonen@nokia.com>
Use a QTemporaryDir for temporary files, instead of the current working
directory.
Change-Id: Ifeb2944238f785a1f7beb0dc2a7c1e092d121db5
Reviewed-by: Kalle Lehtonen <kalle.ju.lehtonen@nokia.com>
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
This autotest seems to be parallel-safe. It was not marked as such due
to an issue which rarely causes the test to hang on exit on Windows, but
that appears unrelated to whether or not the test is run in parallel.
Change-Id: I30bac75be3ddc14139594605481eb6af3f6795e7
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
This autotest fails a parallel-stress test because it writes into its
own source/build directory. However, by inspection, it appears not
likely to cause issues with any tests other than itself.
Change-Id: I13789ba14bab240d34c22c5b77d6407995423afc
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
See the discussion of this topic on the mailing list:
http://lists.qt-project.org/pipermail/development/2012-May/003943.html
The consensus is to not have these methods have inline code to actually
acquire the lock (i.e. no atomic test-and-set or similar). QBasicMutex
is unchanged, and continues to have inlined lock(), tryLock(), and
unlock(). QMutexLocker has been changed to always call QMutex::lock()
(even though the constructor takes a QBasicMutex parameter).
Change-Id: Ic7d2d9d581e6b254c84fdfdd8ce6c425535a8078
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Changed one testfunction to use the test's own QTemporaryDir instead of
the system-wide temporary directory.
Change-Id: I6740a7f4ba7f53174cd0730239d8dc088e5111ba
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Ensure we always use a unique filename when writing to test files.
The test already contained code for this, but it was not applied in a
couple of places.
Change-Id: I1e29ee162c390e014688ab46e3658e2a463d203e
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
In Qt4 QPrinterInfo::defaultPrinter() has inconsistant behaviour
depending on the platform. On Mac and Unix if no default printer
is set in the system then the first available printer is returned.
but on Windows it returns a null QPrinterInfo.
Currently Qt5 returns the first printer on all platforms, but this
causes an inconsistency if you call isDefault() on the returned
QPrinterInfo.
The apidox is slightly ambiguous, but does seem to imply no printer
will be returned if no default is set.
The only place in Qt the public api is used is in QPrinter which is
coded expecting that defaultPrinter() may return null but for there
still to be availablePrinters().
Change-Id: I1bbef8cba259b7d51980a0199e9fd6d5466d921f
Reviewed-by: Teemu Katajisto <teemu.katajisto@digia.com>
Reviewed-by: John Layt <jlayt@kde.org>
Fix an assert that wasn't changed when the "Print to PS" option was
removed
Change-Id: Ife7350976707059861660a74dc2f89d302d861f0
Reviewed-by: Teemu Katajisto <teemu.katajisto@digia.com>
Reviewed-by: John Layt <jlayt@kde.org>
In QTreeViewPrivate::adjustViewOptionsForIndex() wrong index had been
used when referencing to array of viewItems. Variable row is set to the
index of the QModelIndex, however it is not as same as the index in
viewItems[] when there was hidden item in treeWidget. Index of viewItems[]
should be used here. Unit test is added as well.
Change-Id: Idc7eda979e7d09c5a07bd6dffd92b7abbac10e67
Task-Id: QTBUG-25333
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
These tests have passed a parallel stress test on all three of Linux,
Mac, Windows. Mark them with CONFIG+=parallel_test to allow CI to run
them in parallel, saving time.
Change-Id: I19fd333c3c645a67374ca998f6c8530dd236b0f8
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
tst_qprocess::lockupsInStartDetached sometimes locks up on mac.
Mark this as a known issue.
Task-number: QTBUG-25895
Change-Id: I08b1bcf39f2bf373e74509a06415d9ba514b8993
Reviewed-by: Kalle Lehtonen <kalle.ju.lehtonen@nokia.com>
Type information is kept in QMetaType class. QVariant is delegating
operations, so it is better to use QMetaType directly.
Change-Id: I91209fa1c9dc4303d6bd47c96824d3cd64ce5291
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Commit 02b80ebb52 marked itemAt(x, y) and
itemAt(point) as deprecated, causing a warning when code calls the function.
Unfortunately the itemAt(x, y) overload calls the itemAt(point) overload,
causing a compiler warning even when not using any functions from
qgraphicsscene.h at all.
This commit fixes itemAt(x, y) to not call itemAt(point) but use the
non-deprecated items() function directly.
Change-Id: I9c4d3a070d65f3406cdab00345c56d4fc62e76c1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
