QParsedNumber will need it, as do some others, like QAngle.
Change-Id: Iee09f498d2cdbb28d43b94301328ec7858661ad2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Added QFuture::onCanceled() method, for attaching handlers to be called
when the QFuture gets canceled.
Change-Id: I1f01647d6173ba0c1db6641e14140108b33ac7c4
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Introduced QtFuture::connect(sender, signal) function returning a
QFuture object, which is resolved when the signal is emitted.
Task-number: QTBUG-81589
Change-Id: Idbe301eb247b468b9b34f3470c3359d6a7af2f3a
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
A recent change in IBus made it prefer the WAYLAND_DISPLAY envvar in
order to compose its socket path for Wayland sessions. This is because
DISPLAY is unreliable in those environment: It might not be there, there
might be several displays pointing to the same Xwayland server (as it's
the case in GNOME 3.36), or there might even be multiple Xwayland servers
(eg. to enforce inter-app isolation with X11 apps).
Fixes: QTBUG-82910
Pick-To: 5.15
Change-Id: I4883b5d06863ba284883dd95281bed2ce7203e29
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
- don't refer to the class with template arguments in the body of the class
(incl. one where we got the template args wrong)
- make namespace-level swap a non-member friend
- make default ctor constexpr and non-explicit
(default ctors should never be explicit)
- make ctor from std::nullptr_t constexpr
- remove op= from std::nullptr (will be handled by op=(T*) anyway)
- pass QTaggedPointer by value (it's a Trivial Type, so can be passed in registers)
- fix missing QTypeInfo for non-default Tag types
- remove unused include limits.h
- make qHash-able (why is the tag ignored?)
Change-Id: Idee1d685ac365c988698a8637fd5df3accfc1396
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This value is the equivalent of qt_module.prf's MODULE_INCNAME and can
be used to specify a name for the module's include subdirectory. The
default is Qt<ModuleName>.
The include name is stored in the module's target property
MODULE_INCLUDE_NAME.
Change-Id: Ie6c8f6882ee2c3db78884ae5781593c803be3c05
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
The class is documented as obsolete, and the majority of APIs is marked
as deprecated. In this first phase, remove all explicitly deprecated
APIs and trivial implementations.
The test case is complete removed; what's left when code that uses any
of those deprecated methods is removed is not testing anything
meaningful.
For some methods, there is no practical replacement using QScreen yet,
and QDesktopWidget is still used in QWidget internals. Those require
refactoring to only use QScreen before the rest can be removed.
Change-Id: I8f7c968ec566820077221d37b817843758d51d49
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Rather than have a paletteChanged() signal which can be connected to for
tracking when the application palette has changed, then it is better to
use the event that is sent to all windows and the application itself.
That way it is easy for a window/widget or item that cares about the
change to the application font to catch it in the event() function.
[ChangeLog][QtGui][QGuiApplication] Deprecated paletteChanged() signal
in favor of QEvent::ApplicationPaletteChanged.
Change-Id: I95da211e30590e357007cc14d8ee266baceba7b3
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
migrating backend was done if the QNetworkSession changed. With
QNetworkSession gone this was no longer called from anywhere.
Change-Id: I8c995001f9d4c7ae83446b4d29fa62c954c79889
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
And enable the same default when building Qt itself (it's implicit).
Allow opting out on a target-by-target basis, by using the public
qt_disable_utf8_sources() API call.
Change-Id: Ifc19a744d57b96b1c74a6926a0c6628c2a820464
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
QMake variables are always string lists, but some variables are used to
represent single string values. People tend to fill those variables with
multiple values and expect them to be joined by spaces.
Exhibit A: qtdeclarative/tools/qml/qml.pro
QMAKE_TARGET_DESCRIPTION = QML Runtime
QMake supports this for many values. Do the same for Scope.get_string().
Change-Id: I6af22bc1bfed07a4d8eac94d7315051a5bcb7a4d
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
These functions mess in bad ways with the overload set (they, fatally,
attract char16_t, e.g.).
The no-ascii warning now comes from the QChar(char) constructor, so
it's not lost.
[ChangeLog][QtCore][QString] Also removed op=(char), op+(char,
QString), op+(QString, char).
Change-Id: I1471c9ddb4c6869aff6e527516e2c78b80b7d7d3
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Even in Qt 5, remove() can be passed an alias to *this. In Qt 6, with
the advent of substring sharing, this will become even more
pronounced. Use the same fix as was already used in QString::insert().
Pick-to: 5.15
Change-Id: I1a0d3d99fd7dff6e727661646d2cbfdc94df2682
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
QLabel ensures a text control in its focusInEvent, and from that point
on lays the text out using the control. A label with text control will
place the text so that there is space for a cursor and text selection,
resulting in text shifting when a label receives focus without having
had a text control before.
Instead of creating the control only on focusIn, ensure that it's
created for any QLabel that can receive focus.
Change-Id: I26c9df3affa08f2360ad4b94de43bd85e2e2e164
Fixes: QTBUG-84080
Pick-to: 5.15
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
It might happen that the resulting subpath generated by the QDashStroker
will have no lines at all. It happens when further LineToElement items
compare as equal to the starting MoveToElement when compared in 'float'
space. In such case QTriangulatingStroker::process() skips LineTo elements
and the vertexBuffer ends up with unconnected vertices, which generate
unwanted triangles.
See related bugs:
https://bugs.kde.org/show_bug.cgi?id=419240https://bugs.kde.org/show_bug.cgi?id=413220
Change-Id: Ie70955287da47e9f7aa9cdeaea506e1c817b2317
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
As best as I can guess, this used the QString::operator=(char), which
I locally removed. Before that lands in Qt, remove this ... wtf?
Pick-to: 5.15
Change-Id: Ie083fe69500d6b5b633416f89f5dd1d7068c20b2
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Calling linear-complexity remove(int, int) in a loop is O(N²).
Fix by implementing a remove_if()-like algorithm which ensures each
character is written at most once, which is linear.
[ChangeLog][QtCore][QString] Fixed quadratic worst-case complexity of
remove(QString). The function now has linear complexity in all cases.
Pick-to: 5.15
Change-Id: I12f70fbc83fb5da4a9aae4bd02f525d7657cc940
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Instead of writing the hard-coded sysroot into the .cmake file, write
the variable ${CMAKE_SYSROOT}. This makes it possible to relocate the
build to a machine where the sysroot is different from the build
machine.
Fixes: QTBUG-83335
Change-Id: Iaa69feb9a140b050f6b5547929cc940ee0f039ce
Reviewed-by: Rolf Eike Beer <eb@emlix.com>
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
The CI system can test this, but there's little reason to blow
up every developers compile time by 30%.
Change-Id: I7677b0411103fb1301ad087c884cc9bd22a5aee2
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
We now assume that source code is encoded in UTF-8 by default on
all platforms (and verify this with an auto test). Provide
a CONFIG+=no_utf8_source option for backwards compatibility.
[ChangeLog][qmake] qmake will tell the compiler that source code is
encoded in utf-8 by default. This mainly has an effect on Windows, where
MSVC still assumes source code is encoded in the current ANSI code page.
Use CONFIG+=no_utf8_source to get back the Qt 5 behavior.
Change-Id: I6dcafcaeefdea7d3907ccb723aeb7d23ccc0f04f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Centralize, rather than keeping adding constructors from any
array-like container.
A more robust implementation, likely following the converting
constructor for std::span ([span.cons]), is out of scope for C++17
and will require C++20's ranges and concepts.
[ChangeLog][QtCore][QStringView] QStringView can now be constructed
from any contiguous container, as long as they hold string-like data.
For instance, it's now possible to create a QStringView object
from a std::vector<char16_t>, a QVarLengthArray<ushort> and so on.
Change-Id: I7043eb194f617e98bd1f8af1237777a93a6c5e75
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
It has been the case for both QStringLiteral and QByteArrayLiteral
since Qt 5.0, and Q_ARRAY_LITERAL since Qt 6.0.
Since it's definitely surprising, add a note in the docs, which
is "somehow" consistent with the interpretation of capacity as
the biggest possible size before we reallocate. Since it's 0,
any manipulation of the size will cause a reallocation.
(Alternatively: the capacity() is for how many elements memory was
requested from the free store. No memory was allocated, so 0...)
Task-number: QTBUG-84069
Change-Id: I5c7d21a22d1bd8b8d9b71143e33d537ca0224acd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... without moving the mouse.
This allows to update drop action and cursor.
Task-number: QTBUG-56218
Task-number: QTBUG-82934
Pick-to: 5.15
Change-Id: I8b0ac2a008a9dbcc4c2d6abce282e6f169c2f542
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Add an operator QVariant() to QRegExp to keep things at source compatible
as possible.
Add a hack to QVariant::load/save() to recognize the old typeid
for QRegExp and stream them correctly as long as the streaming operators
for QRegExp are registered.
Also move the datastream test for QRegExp to tst_qregexp, and adjust it to
the qvariant changes.
Change-Id: I120b38a7541b43ec07a21b17f7f35c55f071eb75
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Replacement methods do now exist in QRegExp, or
for QRegularExpression when porting to it.
Remove all autotests associated with the old methods.
Change-Id: I3ff1e0da4b53adb64d5a48a30aecd8b960f5e633
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The prepares for the removal of those methods from QString and
QStringList. The new methods in QRegExp are left as a porting help.
Change-Id: Ieffa33a79caf53b83029e9b070c4eb5cadca1418
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
- don't force the deprecation sentence into a separate memory location
(gives the compiler more leeway in how to lay stuff out)
- split the switch (will be useful when extending)
- fix a spelling mistake in one of the messages
Change-Id: Ied137dc8eee7047177983660e1a6776a0bf46bde
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Move pre/and post condition handling out of the main loop
to make that one as fast as possible.
Remove special handling of a corner case when the input length
is zero, where the utf8 decoder did something else than all
other decoders.
Change-Id: I94992767ea15405b38f7953adadaa6ff98b20b6f
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Just so we can get this cleaned up as well and remove it from
Qt Core.
Change-Id: I2b5b821b039ce2c024ec3cb7338a1a9becdd2157
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
There's no real dependency to QTextCodec in those files anymore.
Change-Id: Ifaf19ab554fd108fa26095db4e2bd4a3e9ea427f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
And refactor the code a bit to not convert to unicode twice and
use the mime database instead of Qt::mightBeRichText().
Change-Id: I56f9a732c8ad593e7f050eaad401be536bdf6f98
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: David Faure <david.faure@kdab.com>
This should all be utf-8 anyway, but right now simply exchange the
text codec with a string converter.
Change-Id: If0a230776824598b6378bb402d692c941e371104
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Use QStringConverter instead to convert HTML to a QString. This limits
the amount of supported encodings to UTF based encodings and Latin1.
This is ok, as anything but utf8 is strongly discouraged by the HTML
spec anyway, and the support we have with this change does cover ~98% of
all real world HTML.
Change-Id: Ia610d327624b083c23d3c604aee70517a4a5eb6a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We already assumed that 8bit data is utf8 encoded in all cases
but for HTML. Handle HTML through QStringDecoder now. This
removes support for encodings other than UTF based one and Latin1.
This is ok, as HTML should nowadays always be encoded in utf8 as
well (anything else is strongly discouraged by the HTML spec). In
addition, utf-8 and latin1 together seem to cover ~98% of all HTML
data.
Change-Id: I7e7165edd38cfac395faf72681e5715b6d014c14
Reviewed-by: David Faure <david.faure@kdab.com>
