Make it possible to test the saving/restoring of directories
which is done by default.
Task-number: QTBUG-28855
Change-Id: Id16010b6aea9cd78b9bebdbe1b148eb7f5141f8b
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
This reverts commit d0804ff2dd.
The commit breaks JSC which is used by QtWebKit, QtScript and
indirectly by QtQuick1, which blocks integration stable->dev.
JSC should be fixed upstream, but until then we need to revert
the change.
Change-Id: I3f7b8be08b68181e08422d2cb00d7cd70a7fc07f
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
The exact same hunk is present 100 lines below.
Change-Id: Ia3d61037b29186368e30f95f4162282d38bca972
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
This allows for example the iOS mkspecs to use xcode-select to get the
path to Xcode's /Developer directory, which is then in turn used to
set up QMAKE_CXX (which is needed by configure).
Coprocess is GNU awk specific, so we're just using getline.
http://www.gnu.org/software/gawk/manual/html_node/Getline_002fPipe.html#Getline_002fPipe
Change-Id: I7e88112bea68f039361d6e96ee581eedf129ab02
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
It's not part of the fallback list we read from the plist file.
Change-Id: Ieaf306d4cd51a6bb6e6f41048876d3e674a4b99b
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Change-Id: Ic6b39825cf349f8ad8a56b1fb5dd3855f8675519
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
It's listed in the 10.8 SDK as CT_AVAILABLE_STARTING( __MAC_10_8, __IPHONE_6_0),
but not in the release notes for 10.8:
http://developer.apple.com/library/mac/#releasenotes/General/APIDiffsMacOSX10_8/CoreText.html
The iPhoneOS 6.0 SDK lists it as CT_AVAILABLE_STARTING( __MAC_NA, __IPHONE_6_0),
which matches the release notes, and some 10.8 systems in the wild apparently
do not have this function declaration in the system headers, so for now we'll
be conservative and not assume it's available.
Change-Id: Idbadda58ea95bfca75458d77cb2799c49fba013a
Reviewed-by: Jens Bache-Wiig <jens.bache-wiig@digia.com>
there is no reason why testlib in particular should have them, while
every other module uses the standard strings.
Change-Id: I9b9e45957dfccbf02939c326dcebf06133098ede
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
it differed from QMAKE_LIBS_OPENGL only for the irix/sco/unixware -cc
specs for not entirely obvious reasons. as all these specs are obsolete,
remove it.
Change-Id: I7d50ffa11ff830371ea52c9ebe25e1f1bc56b307
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is possible with QWeakPointer, so allow it for migrating
code too.
In the process, replace the QPointerBase with a member variable for
simplicity. The functionality of the QPointerBase is replaced
by a TypeSelector template.
Change-Id: I3b4c77bdeda2b863cc33e84a3da8a25bae928c8c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
A new glyph type is added to the glyph caches for ARGB bitmap glyphs,
and the raster and OpenGL paint engines have been modified to support
this glyph type for drawCachedGlyphs().
The CoreText font engine implements support for these glyphs through
the CTFontDrawGlyphs API, since CGContextShowGlyphsWithAdvances does
not handle color glyphs.
Change-Id: Idad9ce75a911cae130d65aebe59142772a16fc12
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
If a popup widget(e.g. a popup menu) has been closed due to a mouse press
event, the mouse press event will be consumed by the popup widget and the
widget under the mouse pointer can't receive the event. This will lead to
confusing behavior such as a push button under the mouse pointer is not
been push down.
This patch replay the mouse press event if a popup widget has been closed.
Change-Id: Id493583dfea9e64ab2964e28d559122c43bbc2a6
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
This avoids warnings in compilers that check for unused variables. They
can't tell that the the variable came from a header.
Change-Id: I1ea5e5bbc76d676fbb561bdc8ae6543e758de90e
Reviewed-by: Andreas Aardal Hanssen <andreas@hanssen.name>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Change-Id: I18697037db742d38874c8a95df12c189ccc51068
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
Reviewed-by: Lorn Potter <lorn.potter@jollamobile.com>
It's similar to Q_UNUSED, but this is to be added in the declaration
Change-Id: I2f664129fb1f34f7913ef371d45c2c0fec958174
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We need to clarify what's the status of the implicit capturing
group #0 in both of this methods. The former doesn't include it,
while the latter does for convenience/consistency in the way
we count the capturing groups.
(Note that this last behavior is actually autotested.)
Change-Id: I2170842c2a6dffa34fa56389ceead61a92c07cd1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
(?J) inside a pattern string can be used to allow or disallow duplicated
capturing group names in the pattern string itself.
Although PCRE supports duplicated names, in Qt we don't yet.
Change-Id: I21cd0c41273cd7ef42870ced3a0fad6ba7035cbc
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Information about the pattern (number of capturing groups, newline
settings, etc.) are grabbed when the pattern is compiled the first time.
Studying (=> optimizing) is always done later, after a certain amount
of usages. In case this ever changes, add an assert.
Besides, we're not grabbing any info that require studying the pattern
first.
Change-Id: Ica15fa21f7bf13213288d7090d3396a89900078e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The current size hint is not correct and the text is truncated
when using prefix/suffix.
Update QCommonStyle::sizeFromContents() to get the button
and frame widths into account for the QSpinBox width.
Update sizeFromContents() in the different styles to be
consistent with the change in QCommonStyle.
Update minimumSizeHint(), calculate it using the prefix and
data range. The SpinBox can shrunk over the suffix if any.
Task-number: QTBUG-28863
Change-Id: Ia742232edf8b11d0283e8136c2818928f8755103
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
QStringRef could have been trivial. The destructor is empty, the copy
constructor copies exactly the members and has an empty body and all the
members are POD. Both functions should be removed or defaulted in Qt 6.
When the destructor is defaulted, we can make the constructor constexpr.
We can't do that now because QStringRef is exported and some nasty
compilers (MSVC) like to export all functions, even inline ones, and
then call them without emitting a local copy.
Change-Id: Ie7509fd1a3f737a6c9156ea078d13bb347fc6be0
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
On Mac OS 10.8 and iOS 6.0 we can use CTFontCopyDefaultCascadeListForLanguages
to get the list of fallback fonts, which is preferable to reading the plist
file from the filesystem. The latter doesn't work (is not allowed) on iOS
in any case, so for iOS < 6.0 we use a static list of fallback fonts.
Change-Id: Ibb5e1b4dedd6bfb936f66b53a20f7ced83536ed7
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@digia.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
In some situation (such as iOS currently) we may end up with an empty
fallback list, and we don't want to re-populate the font database
on every call to fallbacksForFamily().
We do not guard populateFontDatabase(), since it's called both initially
and every time the font database has been invalidated.
Change-Id: Ief1342c40f75e5d393e054e9a20bc94bc357d482
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
This finally makes it possible to make windows fullscreen etc from
QML by doing "visibility: Window.FullScreen". I don't see any reason
from not having the API at the QWindow-level instead of at the
QQuickWindow-level since this way it can benefit other use cases too.
Change-Id: If27344306eb563bc2ccd83296a46b1f2862e2db1
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Additions to make the convenience classes work for the
Qt for Android port.
Change-Id: I25ba0faf93c7e09ab04a3fa0784e04631e5ab036
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
This patch changes QGraphicsViewPrivate::updateRubberBand to return
at once in case some of the first conditions (which were needed to
do something with the rubberband) are not true.
The indentation after these ifs is fixed, and there are a few style
fixes, but beside the first 2 ifs, there are only white-space
and new-line changes.
Change-Id: I7e2edb7bfd334b35ee8ab246f733d854bff7e0f7
Reviewed-by: Andreas Aardal Hanssen <andreas@hanssen.name>
Reviewed-by: Andy Shaw <andy.shaw@digia.com>
VxWorks has defined variables with same name as in Qt's headers.
Remove these undefines because that has already done in file
src/corelib/kernel/qfunctions_vxworks.h. See related change
f7bd8652ca.
Change-Id: I994ac9d00ca223b9fa955dfcba8ad6c8dcbd0549
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Vxworks DKM mode does not have rand_r function, use function
implementation from qfunctions_vxworks.h/cpp instead.
Change-Id: I8f23c3453ab9f31280eb90f66dd83d7a64ee98c9
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
It was only used in the now removed OpenGL 1.x paint engine.
Change-Id: I2237172de700bfd31ca25279fbed21d601785962
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Currently, using QDate::maxJd() in tests will fail. This patch changes
some ints to qint64s to prevent overflows where necessary.
Change-Id: I61ebf8f233411a7544689fd5bfa9c3abee54e933
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QTextCodec::codecForHtml currently fails to detect the charset for this
HTML:
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=9,chrome=1">
<title>Test</title>
</head>
This patch makes the detection of charsets more flexible, allowing for
the use of the HTML 5 charset attribute as well more terminator characters
("'", and ">").
I also added a *_data function for the unit tests.
Task-number: QTBUG-5451
Change-Id: I69fe4a04582f0d845cbbe9140a86a950fb7dc861
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Denis Dzyubenko <denis@ddenis.info>
ICC thinks that the return statement is missing:
src/plugins/accessible/widgets/itemviews.cpp(867): error #1011: missing return statement at end of non-void function "QAccessibleTableHeaderCell::parent"
Clearly it's wrong, but the function is overly complex for no good
reason. So just simplify it instead.
Change-Id: Ic63029df18b4712d9864a4dc66b6751b2d4574b7
Reviewed-by: Jan Arve Sæther <jan-arve.saether@digia.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
That's what the macro did anyway, except it had an extra "static"
keyword. This allows us now to use Q_ALWAYS_INLINE in non-static
contexts, if we wanted to (e.g., certain template cases).
Change-Id: I899d359f46234e93d1912ef9125080dc84e15c93
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
... like e.g. in QAbstractSocket::connectToHost().
We can set the scheme on the query URL based on well-known ports; even
if the port is not well-known, we can just query the URL anyhow and let
the netstatus API resolve the right proxy for us.
In addition, return early for local schemes like "file" and "qrc".
Task-number: QTBUG-29425
Change-Id: I900f1ecbac6dd380bb8a77470b39de2c07aa7f6e
Reviewed-by: Andrey Leonov <aleonov@rim.com>
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
Calling popUpMenuPositioningItem:atLocation:inView: will swallow any
mouse up/released event, if we popup on mouse down/pressed. Since we
cache the button state in QNSView, the safest way seems to be to clear
this button state. Another option would be to send Cocoa mouse up
events for each mouse button, but we prefer to keep things simple.
Change-Id: I7f3f369c45e9f5ed9d2ab693bffd8be4a4596ae6
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Change-Id: I491edcae84c149775cf3640626ccff8cbf0d69fe
Reviewed-by: Jon Severinsson <jon@severinsson.net>
Reviewed-by: Andreas Aardal Hanssen <andreas@hanssen.name>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Jason McDonald <macadder1@gmail.com>
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
The compilers checked here are not supported in Qt 5. Additionally,
the QSharedPointer implementation has similar operators without such
guards, so in reality these compilers may not have worked
with Qt 4.6+ either.
Change-Id: I208f3cde7c689770ae15245a555e3a58b749a8a3
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Addon modules may not be in the standard Qt namespace and therefore do
not use the QT_{BEGIN,END}_NAMESPACE macros. Instead module specific
QT_{BEGIN,END}_NAMESPACE_FOO macros are used. Extend the syncqt test
to allow for this and ensure that the QT_END_NAMESPACE suffix matches
the QT_BEGIN_NAMESPACE suffix.
Change-Id: I33110c01118b0fb4376307db7447f747fc38bb29
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
updateRubberBand is only called from mouseMoveEvent.
The event is not changed. At the end of the event
we call d->mouseMoveEventHandler which calls
storeMouseEvent. There is no need to call it twice.
Beside that this patch also changes the argument
from QMouseEvent* to const QMouseEvent*. The only
reason it needed to be const was the call to
storeMouseEvent.
Change-Id: I01148a6a8d36e782c4d95a90e36435e20b1681e1
Reviewed-by: Andreas Aardal Hanssen <andreas@hanssen.name>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>