QFuture's original design pre-dates C++11 and its
introduction of move semantics. QFuture is documented
as requiring copy-constructible classes and uses copy
operations for results (which in Qt's universe in general
is relatively cheap, due to the use of COW/data sharing).
QFuture::result(), QFuture::results(), QFuture::resultAt()
return copies. Now that the year is 2020, it makes some
sense to add support for move semantics and, in particular,
move-only types, like std::unique_ptr (that cannot be
obtained from QFuture using result etc.). Taking a result
or results from a QFuture renders it invalid. This patch
adds QFuture<T>::takeResults(), takeResult() and isValid().
'Taking' functions are 'enabled_if' for non-void types only
to improve the compiler's diagnostic (which would otherwise
spit some semi-articulate diagnostic).
As a bonus a bug was found in the pre-existing code (after
initially copy and pasted into the new function) - the one
where we incorrectly report ready results in (rather obscure)
filter mode.
Fixes: QTBUG-81941
Fixes: QTBUG-83182
Change-Id: I8ccdfc50aa310a3a79eef2cdc55f5ea210f889c3
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Qt6 is binary incompatible with Qt5, no need to keep the old data.
Change-Id: I7fc4592a79ed0a8b79569926a31ef6deb5d3f983
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The function qt_generate_module_pri_file did not take into account the
CONFIG_MODULE_NAME argument of qt_add_module.
This resulted in wrong file names and content, e.g. qt_lib_test.pri
instead of qt_lib_testlib.pri.
Fixes: QTBUG-83176
Change-Id: Id6991396cf9ea5a1d155aa15402c0d84a8a9d082
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Leander Beernaert <leander.beernaert@qt.io>
When the Wayland client plugin is in use, the capabilities
for the surface may report a minimum image count of 4. The
internal "maximum minimum count" of 3, was arbitrary and
only used for sizing buffers. To be more friendly to different
setups, we remove the restriction and use QVarLengthArrays
instead.
We also set the initial size of the buffers to 4 so that we
can run with Wayland without any resizing, but now the arrays
will also grow to be safe for cases where 4 is not sufficient.
Change-Id: Iba5434e84417d36b70f2655b152e816f04650ce4
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
In 519ea72108, the minimum SDK
version was moved away from the AndroidManifest.xml, and when
the change was merged, we lost parts of the update to set
the minimum version to 23.
This redoes it in the remaining locations.
Change-Id: Iada8188d3c0c8ec0a4801c1d219640ecb3976753
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
(This reintroduces c17a5cec19,
which was reverted in Qt 5 because it requires API level 23.)
This flag tells the app to use as much of the screen as possible
while still keeping system UI visible, and can be supported on Android
by using translucent system UI, similar to iOS.
What this does:
1. It changes the current fullscreen/not-fullscreen logic to
allow three states: fullscreen, fullscreen with translucent
decorations and not-fullscreen.
2. In order for it to work, we have to send the actual
screen geometry and available geometry, at least in the case
where the user needs to know the available geometry to
know the safe area of the window. So we get the real screen
metrics and pass these to the QPA plugin (API level 17, so
we can do that now that the minimum version is 21.)
3. Note that getting the insets and calculating the useable
area does not work for non-fullscreen windows, since Android
is quite inconsistent in this respect. So in this case we
just use the window size and origin of 0,0 for the available
geometry.
4. Since we are touching this code anyway, this patch also tries to use
more consistent wording (calling it "available geometry" everywhere
instead of desktop geometry in some places and just geometry in
others, etc.)
[ChangeLog][Android] Qt::MaximizeUsingFullscreenGeometryHint window
flag is now supported, and will make the window fullscreen, but keep
the system UI on-screen, with a translucent background color.
Fixes: QTBUG-74202
Change-Id: I4cc5ef9cc2a3bd22d4d8d2bb767c6ff8a3aa75c0
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
Reviewed-by: Christian Strømme <christian.stromme@qt.io>
- Moves QUndo* classes (except QUndoView) from src/widgets/utils to src/gui/utils
- Moves related auto tests from widgets to gui
- Replaces QUndoAction with lambdas that do text prefixing
[ChangeLog][Undo Framework] QUndo* classes (except QUndoView) were moved from Qt
Widgets to Qt GUI.
Done-with: volker.hilsheimer@qt.io
Fixes: QTBUG-40040
Change-Id: I3bd8d4d32c64f8dee548f62159a1df2126da89d8
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
and replace with QRegularExpression
Change-Id: Ic692fc0ea24da84dd4b6bb4c8a846c0fcc62c3cb
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Samuel Gaist <samuel.gaist@idiap.ch>
If recent versions of libc++ are built with filesystem support disabled,
the filesystem header still is installed., The std::filesystem::path()
constructor is completely defined inline in that header, making the test
pass on such configurations, despite C++17 filesystem not being
implemented.
Test a call to std::filesystem::copy instead, which requires the actual
library implementation to be available as well (on recent libc++).
Change-Id: Id997ab75f3299d8431b13cad871f2901f4d9f6ed
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Remove usages of Qt::AA_UseHighDpiPixmaps from examples/.
This flag is now on by default, and can't be disabled.
Task-number: QTBUG-83092
Change-Id: Ie28622f816da2fe28d4ab272d45ea20bd051d5f3
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Explicitly use the Qt APIs that return QPixmap and QBitmap by value, and
fix the API taking those to use const references rather than pointers
or const values.
Change-Id: I2bb7ad1edb3b65f806f0475fca383e5b9bdb61f3
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
QShortcut has only one widget specific feature, which is whatsThis; that
is just a QString, so the setters and getters can just as well be in
QtGui.
The widgets specific implementation of shortcut matching and of showing
the whatsThis balloon stays in QtWidgets, in the private implementation.
Using virtual functions in the private we can override the empty default
in QtGui, and by adding a virtual factory function in QGuiApplication,
the correct private is instantiated depending on the kind of application
running.
Change-Id: I09ae4a5482f9fb70940c5e2bfe76d3d7fd710afc
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Duplicating the number of classes is a high price to pay to be able to
have some QAction functionality behave differently, or be only available
in widgets applications.
Instead, declare the entire API in QtGui in QAction* classes, and
delegate the implementation of QtWidgets specific functionality to
the private. The creation of the private is then delegated to the
Q(Gui)ApplicationPrivate instance through a virtual factory function.
Change some public APIs that are primarily useful for specialized tools
such as Designer to operate on QObject* rather than QWidget*. APIs that
depend on QtWidgets types have been turned into inline template
functions, so that they are instantiated only at the caller side, where
we can expect the respective types to be fully defined. This way, we
only need to forward declare a few classes in the header, and don't
need to generate any additional code for e.g. language bindings.
Change-Id: Id0b27f9187652ec531a2e8b1b9837e82dc81625c
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
as we raised the minimum requirement for harfbuzz to version 2.6.0:
* use new HB_SCRIPT values instead of tags
* get rid of deprecated (and no-more-used) callbacks
* replace deprecated hb_ot_tags_from_script() usage with a more
flexible and up-to-date hb_ot_tags_from_script_and_language()
Change-Id: I0eafdd2d2028c353fa3a93f5868efceccd364a70
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
When topLeft and bottomRight are different in QAIV::dataChanged(), the
current implementation simply calls QWidget::update() without checking
if the affected cells are visible. This results in a big performance hit
when cells are updated frequently.
Now try to compute the exact update rect by iterating through the
modified indexes.
Fixes: QTBUG-58580
Change-Id: I97de567d494e40ed8cdb1ea1f5b3cf3a2f60455e
Reviewed-by: Samuel Gaist <samuel.gaist@idiap.ch>
This reverts commit cf000d080c.
Reason for revert: This API is not needed (anymore) by qtdeclarative.
Change-Id: I8af604c2babe3afc11f183ddb3ce3a9038a456ad
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
QtQml needs the private just for one detail which nobody else should
need it for: Tracking additional dependencies and marking the binding as
dirty. Exporting the private requires hiding some variables and
providing accessors, to compile with MSVC - including the removal of
QVarLengthArray usage. Upside: The binding structure shrinks by 8 bytes
and the encapsulation makes it a little easier to change things without
breaking declarative, ... in the unlikely event ;-)
Also remove setDirty() from the public API as it's not needed by QtQml
and using it is dangerous, because it means that there's a risk of
somebody keeping a reference (count) to the untyped binding from within
the binding closure, which introduces a memory leak.
Change-Id: I43bd56f4bdf218efb54fa23e2d627ad3acfafeb5
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Speed up the QSortFilterProxyModel filtering by only updating the
source_to_proxy entries which are really changed - When proxy intervals
are added or removed, it is not needed to update the proxy_to_source
indexes which were not touched.
Change-Id: I35459ff1b04f4610ec74f4b01d58a71832a9ae22
Reviewed-by: David Faure <david.faure@kdab.com>
When cross-compiling with CMake, before this patch pkgconfig calls would
find libraries which are part of the host system and not the target
system.
The current approach used is based of the discussion present in
https://gitlab.kitware.com/cmake/cmake/-/merge_requests/4478, and should
be considered a temporary solution until the issue is properly addressed
in upstream CMake.
Change-Id: I535d4d48c2a5d34689082b80501b3b6ae30d7845
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
This header only covers a part of what qsimd_p.h does, namely
the compile time detection of simd extensions.
Change-Id: I05f1d987f194a5bec335f2405cc2846fbaa88b66
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It's not used in tools at all and fits a lot better in global.
Also fix the qsimd_x86* files to have a proper copyright
header.
Change-Id: Id3d8e7cfcd7769a1ca9f3d8cf6d357a31a99ba40
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Those methods don't modify the CBOR value, they should be const.
Change-Id: I213984ed82e43429413344999a22ad1c4c3beac8
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This patch silences the exception about missing v1 api entry for the
qt_add_3rdparty_library() function.
Change-Id: I74a473ed1c063b13f0a9ca64dcb568b8f8b27235
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Oh, and re-remove the spdy protocol handler.
Change-Id: I9ca3517b14ac202eac19ae75d481cfd78d431abb
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
With very complex regular expressions this makes a significant
difference in performance.
Change-Id: I5267bf71cae98dffb58ec7a1fc982dae639ff312
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
hash.erase(hash.constFind()) is bound to crash if the hash doesn't
contain the item we're looking for.
Change-Id: Icbefca87b0258970373ec55d5dc113e6ab39c5f0
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
For INSTALL_*DIR variables that have the the same value as
CMAKE_INSTALL_PREFIX, a second cmake run cleared the value. This is
because file(RELATIVE_PATH) returns the empty string if we pass the same
absolute paths.
Fix this by checking the return value of file(RELATIVE_PATH) for the
empty string and setting it to ".".
It's a limitation of qmake that empty strings are not handled as ".".
Change-Id: I8fc4d1eabcc9d5634be2f3741b0002a347dd17e6
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Without this patch JPEG support is always disabled, even though we can
build it from 3rdparty sources.
Change-Id: I9e619f0ca8ec3ca3e7c58981bb6af9b33426a029
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
PCH files were only used while building qtbase. Make sure the value
is exported to the BuildInternalsConfig file, so the value is re-used
when building other repositories.
Also disable PCH when building simulator_and_device iOS builds, because
CMake doesn't currently generate separate PCH files per architecture.
Change-Id: I79955ebc557b800bc3c704deac519fe80012c229
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
mappedObject() can do exactly the same thing, there's no need for a
mappedWidget() signal as well
This removes a reverse dependency between Qt Widgets and Qt Core where
QWidget pointers are being used inside Qt Core.
[ChangeLog][QtCore] Removed QSignalMapper::mappedWidget. Connect to
mappedObject instead, and use qobject_cast<QWidget *> to handle Widgets.
Change-Id: I60618a34dd94575e993bd6d8a2cead95cfd71d55
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
As a drive-by, enable testing of rcc in the cmake build.
Change-Id: I4150ff3ffe7404bab0cbc72f80b23b47a60cf33d
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: Samuel Gaist <samuel.gaist@idiap.ch>
