Change-Id: Ic578e9c3f4d8f70bcac0d95f59942cd2724aee51
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
As a matter of fact also reduces state duplication with QQnxWindow
by getting rid of m_surfaceSize.
Change-Id: I47f5af0f00a317d3bb246c0c9161e2219fbd3263
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
Some QWindow subclass rely on those events to be sent to know when
the window geometry changes (e.g. QQuickCanvas).
Change-Id: I16d3928803d09761f265544fdf10a5d080fccc17
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
Change-Id: I82e464df86b55f04498a8f6f5c1cc5011cac0bd6
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
Maximized/fullscreen windows geometry should follow screen geometry,
so provide a convenience method implementing this behavior. It is
meant to be used in platform plugins where the platform doesn't
automatically resize said windows on screen geometry changes.
Change-Id: Id9128fee1ddf587a7d96aa294d2d1e6eb6d6d11b
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
As QTBUG-25076 points out, the ### comment indicates
that we want to add virtual function to
QMacPasteboardMime.
The default implementation returns 1. Reimplement
for QMacPastebardMimeUrl and return the url count.
Change-Id: Ie300574eab9991af625986805d2b030914291cc0
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
This macro is now left defined from qglobal.h, so we don't need to
redefine it everywhere (and risk getting it wrong).
Change-Id: I2a11d10fe0434b85e79d0dda5f11fa90e2edc431
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: Denis Dzyubenko <denis.dzyubenko@nokia.com>
QFactoryLoader::indexOf() expects the meta-data to contain a Keys
entry, if there were any keys. For compat plugins, the result of
the plugin's keys() function should be stored here.
Change-Id: Ifd04f90cbfce2598d71548c469baa55ca6b0b338
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Change the way the printsupport plugin creates QPrinterInfo
objects, provide platform api to return a named printer, and
expose this as static public api in QPrinterInfo.
Only the Mac plugin used the old api, the other plugins will
have direct support added in separate commits, but will use
the default implementation for now.
Change-Id: I7d6b6556eb39919cfb15bc0e814afbaf13c5712c
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Rename QUDeviceHelper to QDeviceDiscovery and add a static device
discovery fallback in case we dont have udev. The fallback so far only
scans /dev/input/event* and /dev/dri/card* at startup and detects device
nodes only by device path.
Change-Id: I7a423910b30ae16a10d8f1f47b86c6b4d2c2ec36
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Reviewed-by: Donald Carr <donald.carr@nokia.com>
directfbegl requires libQtPlatformSupport to provide some EGL
helper functions. Enable EGL by adding it to the QT_CONFIG variable
in the qmake.conf.
Change-Id: I80d3163d9da4b9eb232aea72ba7d4388453f8247
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
When the QNetworkAccessHttpBackend was merged into the
QNetworkReplyHttpImpl there was some code and comments comming from the
backend that was left in the new class. This removes some of these
leftovers from the QNetworkAccessHttpBackend.
Change-Id: Ifa118160438e2740fb9bf52907066096d8de9ae7
Reviewed-by: Markus Goetz <markus@woboq.com>
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
MSVC2010 32-bit (with and without service pack 1) takes about 1 hour to
compile this file in some builds, since
1c7421ad14.
Avoid the relevant portion of the code just for these compilers.
Change-Id: Icbb4fa12a6d563a7cdc882c30cdb5705675bedb0
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@nokia.com>
Change-Id: I4a7c033691ed83698ac161c034f795f1ac3b0ade
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Reviewed-by: Andreas Holzammer <andreas.holzammer@kdab.com>
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Mac in Qt4 and 5 has been using the PMPrinter Name for the QPrinter
and QPrinterInfo printerName() value, but this is incorrect. This
is in fact the CUPS Description field, is in human readable form
and is not guaranteed to be unique. The CUPS Name field is the
PMPrinter ID value and should be used as the unique identifier
when accessing printers. This has worked up to now due to an
undocumented feature in the OSX api that accepted the Name when
the ID should be used.
Changing all uses of PMPrinterGetName to PMPrinterGetID fixes this
and allows the QPrinterInfo test of names to pass without
munging the names.
Change-Id: I25322aa1a924bed9f67f4ad5e208274c8b700e17
Reviewed-by: Lars Knoll <lars.knoll@nokia.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
Reviewed-by: John Layt <jlayt@kde.org>
Change-Id: I4dc13e5388e065c2425641c7efabcdee8f94800e
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
The file was added with 'x' bit incorrectly in 1f55af8.
It was then changed to the correct mode in 6af65edcb.
But it got back the 'x' bit again in 98dbaec.
Change-Id: I55f267935c17e133ad593f027f795a9ec381fcaa
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Re-enable checking if the compiler is in the path. The previous
commit dealt with a user/spec author not setting CROSS_COMPILE and
then picking up the host g++. Re-add the 'which' check, but put it
after the sanity check for the CROSS_COMPILE variable. This check
assumes that QMAKE_CXX is of the form "${CROSS_COMPILE}g++".
Change-Id: I54f7e058a75d26d73eca5a860946a6854ce91d67
Reviewed-by: Girish Ramakrishnan <girish.1.ramakrishnan@nokia.com>
Calling handleMouseEvent() with w == 0 implies that the local position
is bogus and instead it should be calculated from the global position
once the target window is known.
Change-Id: If173d0570f6dcc8b7bc5d6f21fa1f69d06d9d702
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Reviewed-by: Paul Olav Tvete <paul.tvete@nokia.com>
The new randomization of QHash is enabled by default. There may be cases
where you need deterministic behavior, e.g. for debugging or regression
testing. This patch disables randomization if QT_HASH_SEED is defined.
Change-Id: Idfad55ea7aba830add0a36334f0f763c62fdce13
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Single signon code path gets the NTLM responses from the system,
so we can't predict the contents.
Task-number: QTBUG-25851
Change-Id: Ia8aa1741ae5af9e48643331bf9a3768550a30166
Reviewed-by: Martin Petersson <Martin.Petersson@nokia.com>
Check for blacklisting in case the application has blacklisted
a cert before windows has (currently unlikely as the blacklist is
hardcoded in Qt)
Don't need to check for time validity because that's already checked
by the windows API.
Change-Id: I34da5c4a8a0f8851b9b7668fc421a93c360c8588
Reviewed-by: Richard J. Moore <rich@kde.org>
QTBUG-4031 mentions that canonicalFilePath() will cause the operator
to return true for the following comparison because it returns an
empty string if the path is empty or non-existant:
QFileInfo("") == QFileInfo("non_existant_file")
I'm assuming that the reason for not checking whether one of the files
exist is based on performance, and so I've updated the comments for the
operator to explicitly state that the result of the above comparison is
undefined.
Task-number: QTBUG-4031
Change-Id: I9b34f189f1628f9362b3604445706abd2342fd6e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QTBUG-4031 and QTBUG-4036 mention that QFileInfo::operator==
includes a size() comparison as part of its equality check. I've
removed this check as it doesn't seem to be integral to the comparison.
Task-number: QTBUG-4031
Task-number: QTBUG-4036
Change-Id: I5663ec0e1ac8f70e0a156357c284696779ecd380
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It needs a 2.0 gamma (as in Qt 4) in platform plugin styleHint.
Change-Id: I2286532c607554c41baee4fa348cd7899bf010e1
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
When given an invalid url, the output shouldn't be a valid url.
KDE's kurltest detected this regression compared to Qt4, where
all invalid urls were empty in toString() -- but we don't want that,
to give as much feedback as possible to the user.
Change-Id: Ie53e6e1c0a1d4bb9e12b820220dfb7e2f7753959
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This causes a crash when using QGraphicsWebView with a GL viewport.
Change-Id: I64bdcfe8cd4d8ed7794188c1df08f9028479f8d4
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
Some keyboard layouts have physical dead keys (like the ¨ key on a
Norwegian keyboard). These do not send any text, so we should not use
[NSString characterAtIndex:0] if the string is empty. When encountering
an empty [NSEvent character] string, use Qt::Key_unknown and
QChar::ReplacementCharacter.
Change-Id: I7281aa9ea6005341c0dcfa5900bfe601e4eac6a9
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
When coverting NSEvent characters to a QString, use
QCFString::toQString(), which copies the UTF-16 characters out of the
NSString into a QString (instead of converting to UTF-8 and back again
on each key press).
Change-Id: I78f7c46e75283b7b51c762fd7f22e351ac068072
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
When receiving keyDown events on the Mac, we pass them to NSView's
interpretKeyEvents method, which ends up calling our insertText method.
This means we end up sending input method events for most text input
(instead of normal key events). While this is not wrong, it is
unexpected, and different form how Qt 4 works.
Change insertText to do nothing if no preedit string has been set when
handling a key event. For normal text input that does not need input
method handling, we can simply return and let handleKeyEvent send key
events instead.
Our insertText implementation has to take care to not ignore calls from
outside key event processing (such as the user input characters with the
Character Viewer). As a result m_keyEventsAccepted changes to
m_sendKeyEvent to let insertText know whether or not insertText was
called via handleKeyEvent.
Change-Id: Ie8c2eb67f6a8438f08442f9ff48f2027a041ca23
Reviewed-by: Morten Johan Sørvig <morten.sorvig@nokia.com>
The number of glyphs returned should take surrogate pairs into account.
The glyphs array and advances array as well. This follows the approach
in QFontEngineFT in general.
Change-Id: Ic53faa5e38c2219b987d76aec434558dad92015a
Reviewed-by: Zeno Albisser <zeno.albisser@nokia.com>
Calling a static function non statically causes an unused variable
warning for the optimised away d pointer.
sscanf causes an insecure functions warning. (Even though it was used
safely in this case)
Change-Id: I07700e2155284ef3ebbe7d604ed59b2e61ee7f95
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Change-Id: I777ebd2c1ca51d244de714eb5b235d931df044d6
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Samuel Rødal <samuel.rodal@nokia.com>
e.g. when a global-static configuration object uses QTemporaryFile
for saving to disk (via QSaveFile).
Change-Id: I532aec6de7411e10e4461d5a4ac60686e2f482df
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Make the QPlatformMenu* classes abstract with
pure virtual functions. Delete qplatformmenu_qpa.cpp.
QtGui no longer has to carry the empty implementations.
Change the inheritance tree - Make QPlatformMenu
and QPlatformMenuBar inherit QObject instead of
QMenuItem. Add (now) missing virtual functions to
QPlatformMenu. The effect of this change is to
reduce the number of pure virtuals the platform
has to implement.
Change QPlatformMenuBar::syncMenu to take a
QPlatformMenu pointer instead of a QPlatformMenuItem*
pointer.
Fix a couple of header/unused variable
compiler warnings.
Change-Id: I77450e40c6aae26f3e1aca204603acebd242c036
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@nokia.com>
Reviewed-by: Bradley T. Hughes <bradley.hughes@nokia.com>
These tests have failed a parallel stress test and may contribute to
instability in test runs.
Change-Id: Ibbbe01f7d9550b953fc9fbd6ed52fc99fdb5f5d7
Reviewed-by: Toby Tomkins <toby.tomkins@nokia.com>
Reviewed-by: Kalle Lehtonen <kalle.ju.lehtonen@nokia.com>
This prevents a possible race condition in which the nvaigator service
notices that we have created a window and starts sending us messages
possibly before we have registered with bps for navigator messages.
This resulted in applications sometimes missing navigator messages at
startup.
Change-Id: I362fd782d7798990bc196f23e07b58293b13275d
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Nicolas Arnaud-Cormos <nicolas@kdab.com>
It should return QWidget::inputMethodHints() instead of QVariant()
Change-Id: I01f5de8f2087ac67d125f54f08abed523653eb92
Reviewed-by: Joona Petrell <joona.t.petrell@nokia.com>
Qt's configure allows for spaces between an option and its
argument. This patch brings -l in line with -L et al.
Change-Id: Iea5c78dc06f69c21b7419fd9fa73c52ae1cea18f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
Added #include <unistd.h> to tests/auto/network-settings.h,
so qtbase auto tests successfully build.
It is needed after the header dependency changes, part of gcc 4.7.
Change-Id: I76d1082f8454263f2c22c31a13aa3c1bf6a0c82f
Reviewed-by: Shane Kearns <shane.kearns@accenture.com>
