[ChangeLog][QtCore][QJsonDocument] Fixed a bug that caused
QJsonDocument's equality operator to crash if one of the operands was
default-constructed and the other wasn't.
Pick-to: 5.15
Fixes: QTBUG-85969
Change-Id: I5e00996d7f4b4a10bc98fffd1629f835f570ef6b
Reviewed-by: Sona Kurazyan <sona.kurazyan@qt.io>
This follows up on commit 580e9eedf783ccbdcb67baa3d1a9dcdd53922f86; if
the allowed range is narrower than some time interval whose end it
straddles, users can only access the later part of the range if
keyboard tracking is disabled.
Pick-to: 5.15
Task-number: QTBUG-65
Change-Id: Ie8882bd65bda603133abaa82ad83eb1cdd91f175
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
QIODevice::skip() called a virtual QIODevicePrivate::skip() to
implement an efficient skipping on I/O devices for the internal
subclasses. The user subclasses cannot inherit QIODevicePrivate, so
this functionality was not externally accessible.
This patch replaces QIODevicePrivate::skip() with a virtual protected
QIODevice::skipData(). While the basic implementation simply discards
the data by reading into a dummy buffer, users can reimplement this
function to improve the performance in their subclasses.
[ChangeLog][QtCore][QIODevice] Added virtual protected skipData().
Now, subclasses can implement device-specific skipping of data.
Change-Id: I9522f7f7ab9d03ac06e972a525f8ec2fa909a617
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
We only support version 4 and 5 in Qt 6.0. 1 and 2 are already gone
(due to being based on binary JSON), now we remove 3 as well.
Task-number: QTBUG-81346
Change-Id: I3627dcc0587f1e36f11e93edf7172889e911d64e
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
Use the actual minimum value, rather than minus the maximum, and adapt
the overflow expectations so that they're correct west of Greenwich as
well as east.
Change-Id: I7a5f4510db0fdea3855b5b2bd4c4a86882030efd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Due to a limitation of mktime(), we would have declared it invalid.
Tidied up qt_mktime() slightly in the process.
Change-Id: I25469e314afee6e0394e564bc69a98883005d4ec
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Petty tidy-up, narrowing the scopes and asserting that localtime_r()'s
return, when non-null, is the pointer we gave it.
Change-Id: I6c0959524260028ca9b234f6d33eae78f27c1412
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The ResizeObserver API is now available on the major
desktop browsers (Safari, Chrome, Firefox), and can
be used to handle canvas resize.
(Previously, we got a callback on viewport resize
only)
Pick-to: 5.15
Change-Id: I8737285416bef70641f90da793c85efcb24f3623
Reviewed-by: Lorn Potter <lorn.potter@gmail.com>
For long double: they're just missing, so add them for completeness.
For integral types: follow the advice of C11's trigonometric
functions; first convert the angle to double, then do the actual
conversion. This is offered only for the degree->radians conversions,
as someone may legitimately want to call e.g. qDegreesToRadians(90).
On the other hand, it seems extremely unlikely that someone may
want to do a radians->degree conversion starting from integral
datatypes, so I'm not adding it for the moment being (instead,
I'm leaving a note).
[ChangeLog][QtCore][QtMath] qDegreesToRadians now also accepts
long double and integral types. A value of integral type will be
casted to double before the conversion to radians.
[ChangeLog][QtCore][QtMath] qRadiansToDegrees now also accepts
long double.
Change-Id: Ib1576be5193ae09bb6cb4a70d7a31702955df2c5
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Interbase is case sensitive when tables/fields are escaped so we should
ensure that we pass the value as is.
Pick-to: 5.15
Change-Id: Ia6c4edc9c1e675bd95913c85d47bf22c418d2113
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
We can provide those. They don't lose information, do not have the
problem we face at offering ranged constructors (namely the order of
duplicate keys), and have a distinct advantage over ranged constructors:
a non-shared rvalue QMap can be "upgraded" to a QMultiMap without
allocating memory for the multimap.
Change-Id: Ic23c83927c05a210bc1f0050006c9f26365d3916
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Forcing the reading of all file system attributes is generally useful in
multi-threaded applications, and has at least one more use case in Qt.
Using it in QFileSystemWatcher on Windows avoids several file system
accesses.
Task-number: QTBUG-41373
Change-Id: Ib3c3243fd083142393ad46d62f49cb4f7bfda17c
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
We use qt_wrap_cpp instead of AUTOMOC so that we can easily find the
json files.
Getting autorcc to run only after json file has been generated was
deemed too tricky. Therefore the test is slightly modified to check for
the json files in its directory instead of the resource, if the qrc file
does not exist.
Change-Id: Id1aabb117c8bab3ff81156da1f66d64e796bf18b
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
This adds supports for setting and retrieving the selected
MIME and name filter for 'QXdgDesktopPortalFileDialog'.
For preselecting a filter [1]:
> current_filter (sa(us))
>
> Request that this filter be set by default at dialog creation. If
> the filters list is nonempty, it should match a filter in the list
> to set the default filter from the list. Alternatively, it may be
> specified when the list is empty to apply the filter
> unconditionally.
The "current_filter" return value was added in xdg-desktop-portal
commit [2].
[1] https://flatpak.github.io/xdg-desktop-portal/portal-docs.html#gdbus-org.freedesktop.portal.FileChooser
[2] 35fca7fae8
Fixes: QTBUG-85658
Pick-to: 5.15
Change-Id: I8651c1a8942dfd358895b7826730729c4d22e680
Reviewed-by: Jan Grulich <jgrulich@redhat.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Embed package versions into the qt_find_package calls for various
Linux specific packages.
Task-number: QTBUG-82917
Change-Id: I5d1cb623f81932dfae4658b8a3a89eedb71ea3af
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
If certain 3rd party libraries have a version that's not suitable for
Qt, the configure summary should say so, rather than use them and fail
at build time.
With the current situation, we have to duplicate the version
information from the configure.json files in helper.py, by assigning
the version number as an extra find_package variable.
Rerunning configurejson2cmake then embeds this version info into the
qt_find_package calls in configure.cmake.
Some of the Find modules are rewritten to take the specified version
into account when looking for the libraries.
This involves moving around the code for creating a target, after
calling find_package_handle_standard_args() so we know if a good
enough version was found.
Task-number: QTBUG-82917
Change-Id: I139748d8090e0630cda413362760034dc3483e11
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
The comparisons with pointer were ambiguous as the comparisons could
be done between iterators or pointers.
Change-Id: I0484946931502d04bd63519fcd6e886c732758d3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Multiplying 65535 by 65535 would exceed the precision of an int, so
use 65535U instead.
Change-Id: I066e552fb7db03ce867bcbfbd0b555ac98ca4bbf
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It's undefined and causes a division by zero.
Fixes: oss-fuzz-24273
Pick-to: 5.12 5.15
Change-Id: I3d34d5c43cccff9255abaf87af265ddea3fe6d95
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
As instructed in a // ### Qt6 comment
Added missing #include for QDateTime.
Change-Id: Ic1f9e6ec42ecae07d037b84943444785847bdf98
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Since 5.0: set{En,De}codingFunction()
Since 5.13: readLink()
Change-Id: I5386d0accf2724d84550c9bfdbbe914937194be2
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Assignment from QString and addResourceSearchPath(), both deprecated
since 5.13.
Change-Id: I25f08ffadc7b9dfd7895a9199255ca5f1948bd47
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Removed isCompressed(), deprecated since 5.15, and (since 5.13)
addSearchPath() and searchPath().
Change-Id: I4b6fb8077c02bbe322334e474eaf0a2a7caf0004
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
QString's fromUtf16() prefers char16_t data over ushort.
Change-Id: Ib20c5afa09ceabb4e91fe434b6a057edb4739a53
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Follows up on commit 9dc1edb314, giving
the relevant advice on what to use instead and protecting with the
appropriate version-check deprecation macro.
Pick-to: 5.15
Change-Id: I4191493e6c43448c4390bf22be1571611b172950
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The header pulled in qhash.h for no readily apparent reason.
It did so subject to deprecation #if-ery, so rip it out.
Change-Id: I00529dd2b2de11d9a997b6fa766901f5b2f4b254
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Employ RO0; inherit the constructors from the base class; and
do some code tidies as a drive by.
The inherited constructors bring in goodies like initializer_list
support.
Change-Id: Ia00a3f9b0ccbf182bf837bc65ba2305110c8dc60
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Introduce a Q_OFFSETOF() macro that uses the optional support of
offsetof() for non standard layout types and disables the
corresponding compiler warnings. All our supported compilers
support offsetof() on non standard layout types.
Use the macro to do the offset calculations required in moc
generated code to replace a manual offset calculation that
was dereferencing a null pointer.
Change-Id: I4aab3af3c8bbaa90372f2234aa1cf8399d023c22
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This enables QDir and QFileInfo/QFile to work with entries found from
a entryList() as it will obtain the permission for these files correctly
when it is encountered.
Due to what seems to be a quirk in the Android API, we cache whenever we
come across a known directory to save time later on for checking if it
is a directory. All uris accessed where we get permissions for are
cached so we get the right URI for that given content url as some are
granted via the Intent.
Fixes: QTBUG-85538
Fixes: QTBUG-83041
Fixes: QTBUG-76886
Pick-to: 5.15
Change-Id: I685b3c60804812a0e4b85fbdbb4ec5efaa09420c
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
We observe changes to the NSWindow to pick up color space changes,
so we need to track when the platform window is destroyed and created
to match our observer with the recreated NSWindow.
This does not handle the fact that we're internally recreating the
NSWindow if certain window flag changes requires so, without letting
clients know via surface events.
Task-number: QTBUG-85915
Pick-to: 5.15
Change-Id: I7a7d728c742def79adebaadc985cedd86ea0d581
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The observer we add for the NSWindow of the backingstore window
will not be valid once that window is destroyed, but we may get
last minute notifications for it still, in which case our platform
window is gone.
This patch fixes the immediate crash, but reveals a more fundamental
problem of assuming the platform window that's alive during construction
is the one that will always be used with the backingstore. This is not
the case when for example QtWidgets opens a combo box, and reuses the
backingstore for the widget each time the combobox popup is shown.
Fixes: QTBUG-85915
Pick-to: 5.15
Pick-to: 5.12
Change-Id: Ia66a45d003882602ac29476aabf2d58b0ac33c1e
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
This amends the parent change but intended only for dev and not
to be picked for 5.15.
Change-Id: If7a2d81045c0625f19554eaf6b5cf69e72d42384
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Before clang 9, it was enabled by default
when -fsanitize=undefined was set.
Pick-to: 5.15
Change-Id: I0faf3ae1901d4badc6d265fa8081185be4360636
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Currently it is possible to compare a QSslError with a
QSslError::SslError because QSslError has an implicit
constructor. But the comparison often fails because
a QSslError received from the system contains a
certificate.
[ChangeLog][QtNetwork][QSslError] The constructor
QSslError(QSslError::SslError error) is now explicit.
Change-Id: I36cc5895245d3b43ab4b8d65a9635893d6b0e6a4
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The "populated" variable is otherwise never written into.
Change-Id: I979411a19927dc4e7e09c6c36edfb2308f519596
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Most of the time lld just gets stuck (deadlock) waiting on some
mutex, thus failing integrations.
Amends 64c111e10f
Task-number: QTBUG-85911
Change-Id: Id73bf967a7aeb0e0cbccfaafe056b325c8711f82
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
Static Qt plugins should not be included in the list of libraries of
a prl file. They end up being there due to our circular dependency
trick where the plugins depend on the module they belong
to.
This in turn causes the giant static plugin generator expressions to
be processed in qt_collect_libs(), and the generated prl file ends up
having target names like Qt6::QJpegPlugin which are obviously not
linker flags.
To eliminate the static plugins from prl files, add an additional
dummy boolean generator expression '$<BOOL:QT_IS_PLUGIN_GENEX>'
that always evaluates to true. We can string match on this expression
in qt_collect_libs, and thus remove the whole static plugin genex
entry.
This should fix linking of apps with qmake that use a CMake-built
static Qt.
Task-number: QTBUG-85865
Task-number: QTBUG-85240
Task-number: QTBUG-85801
Change-Id: I949dc5edb10d94c4ab84ed430af7c164d8efaaa6
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
In particular switching the SUBPIXELS macro with FRACT makes the logic
valid for all 24dot8 fixed point values.
Change-Id: I4e58df94f8cf3c557f670c5d3088942e9b8efa6d
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
mkspecs/features/qt.prf adds a dependency on the system threading
library if the Qt Core thread feature is enabled. Because qt.prf is
loaded by any public or internal Qt project, it's essentially a public
dependency for any Qt consumer.
To mimic that in CMake, we check if the thread feature is enabled, and
and set the Threads::Threads library as a dependency of Qt6::Platform,
which is a public target used by all Qt modules and plugins and Qt
consumers.
We also need to create a Qt6Dependencies.cmake file so we
find_package(Threads) every time find_package(Qt6) is called.
For the .prl files to be usable, we have to filter out some
CMake implementation specific directory separator tokens
'CMAKE_DIRECTORY_ID_SEP' aka '::@', which are added because we call
target_link_libraries() with a target created in a different scope
(I think).
As a result of this change, we shouldn't have to hardcode
Threads::Threads in other projects, because it's now a global public
dependency.
Task-number: QTBUG-85801
Task-number: QTBUG-85877
Change-Id: Ib5d662c43b28e63f7da49d3bd77d0ad751220b31
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
