376e3bd8ec added the new class for Qt 6.1,
but during header review we concluded that using a class introduces
complexity wrt instance ownership and API design that can be avoided by
using a std::function instead.
The functionality is tied to QTextDocument, so the type definition and
the default provider API is added there.
Since std::function is not trivially copyable, the atomicity of the
previous implementation is not maintained, and concurrent modifications
of and access to the global default provider from multiple threads is
not allowed. The relevant use case can be supported by implementing a
resource provider that is thread safe.
Task-number: QTBUG-90211
Fixes: QTBUG-92208
Pick-to: 6.1
Change-Id: I39215c5e51c7bd27f1dd29e1d9d908aecf754fb7
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
We deliberately expose moc to warning-worthy C++, but we don't want the
compiler warnings coming from those.
Change-Id: I09c36d7c8bfbb88f7bde61fa7bcc52a5adde81da
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
While the class name is now a bit more than a mouthful, it's purpose
is very narrowly tied to QTextDocument, so don't use a very generic
name for it. That resources are provided based on a URL is to some
degree an implementation detail, and URLs are resource locators so
we don't need that in the class name.
Address code review comment for 6.1. Add documentation and links to
existing APIs with a similar purpose.
Task-number: QTBUG-90211
Task-number: QTBUG-92208
Pick-to: 6.1
Change-Id: I4f09057cc2f53a5595513c1c9422e6ccaad6ca13
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Q_CC_GNU is defined on compilers that masquerade as GCC (Clang, ICC),
so using it to work around GCC-specific bugs is wrong. Introduce a
local define for _only_ GCC and use it in place of Q_CC_GNU.
Drive by: version-fence a test we now know it's been fixed upstream,
and correct the link to the corresponding bug report.
Pick-to: 5.15 6.0 6.1
Change-Id: I9059d6e6bf86157aca71590ac22afb1a1c114313
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
fillRule() contained a major blunder: instead of checking if the
d-pointer was allocated, and return a default value if it wasn't,
it checked whether the path contained any points. This means that
QPainterPath p;
p.setFillRule(x);
Q_ASSERT(p.fillRule() == x);
was failing.
As a drive-by to test this change, fix another mistake in clear():
clear is documented to clear the elements in a path, but instead
it also changed the fill rule.
This commit partially reverses 697910e5fb.
Change-Id: Ieb8145694b672439c3380d9ccb87d1206a2dd115
Pick-to: 5.12 5.15 6.0 6.1
Done-with: Milian Wolff
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
This function returns the metatype corresponding to the metaobject, or
an invalid metatype for namespaces.
This works as follows: First we increment the metaobject revision for
new metaobjects. Metaobjects with older revisions are handled by doing a
lookup by name. That fallback is also used for dynamic metaobjects (from
QtDBUS and those created by QMetaObjectBuilder).
For new metaobjects, we store the metatype in its metatype array, behind
the property metatypes. This avoids any changes to the property and
method metatype extraction logic: For properties, the metatype access
does not change, as the new metatype is after their metatypes. For
method metatypes, we already have an indirection layer (using offsets),
so by adjusting those offsets by one, the same logic keeps working.
To distinguish between namespaces and dynamic metaobjects, namespaces
store the metatypeinterface pointer for void in the metatype array,
whereas dynamic metaobjects store a nullptr.
One nice additional benefit is that this simplifies the generator logic
in moc, as the metatype array is now never empty.
Task-number: QTBUG-92077
Change-Id: Id3f920f28553f12032a71a1a87dad29e5374dbe7
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
We are still using Catch2 2.11.3 which is not adapted to Apple
silicon yet. This patch backports the required change
from Catch v3.0.0-preview.3.
Change-Id: Ifa14a1fdd6cd1f661c94a0a78648cb01bd9699c1
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Fix operator checks for containers whose value_type equals themselves.
It does not make sense to recurse on value_type in that case. Thanks to
std::disjunction having short-circuiting semantics, we can avoid that
issue by checking first whether T is T::value_type.
As a drive-by, check for value_type typedef before checking for
begin/end in is_container. This works around an issue in gcc <= 8.1,
which fails to correctly SFINAE the case where begin and end are private
methods.
Pick-to: 6.0 6.1
Fixes: QTBUG-89456
Change-Id: I27305a7cfe050f13a279c07f00bc229c01daa25b
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
It was treated differently depending on format, made it consistently
behave the same for all formats (following the behavior of the primary
formats).
Pick-to: 6.1 6.0 5.15
Change-Id: Ie24e19957d076fdf3ebd333074e26ede187489eb
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
QGridLayout::takeAt() and QLayoutItem *itemAt() only check the upper bound.
If the index < 0, these function will return invalid pointer.
Fixes: QTBUG-91261
Pick-to: 5.15 6.0 6.1
Change-Id: Idfb9fb6228b9707f817353b04974da16205a835c
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
QSqlQuery is a broken value class. Copying one object would mean
copying database state (the result set, the cursor position, etc.)
which isn't generally available for all database drivers.
For that reason, the current implementation does not honor value
semantics -- modifying a QSqlQuery object has visible side effects
on its existing copies (!).
The correct solution is to accept that QSqlQuery is a move only
type, not a value type. Add move semantics to it, and deprecate
its copies.
(We can't just *remove* copies in Qt 6 due to SC/BC constraints).
[ChangeLog][QtSql][QSqlQuery] QSqlQuery copy operations have
been deprecated. QSqlQuery copy semantics cannot be implemented
correctly, as it's not generally possible to copy a result set
of a query when copying the corresponding QSqlQuery object. This
resulted in modifications on a QSqlQuery having visible (and
unintended) side effects on its copies. Instead, treat QSqlQuery
as a move-only type.
Fixes: QTBUG-91766
Change-Id: Iabd3aa605332a5c15c524303418bf17a21ed520b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
This can only be completed in Qt 7, due to BC guarantees. I've only
updated the source code for QHostAddress and its unit test and even then
I did not touch the documentation. This needs to be completed in Qt 7.0
Change-Id: I26b8286f61534f88b649fffd166b695882f8f3b5
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
The calculation rounds early and often, which is intentional. Add
unit test to make sure we don't regress.
Fixes static analzyer report about incorrect rounding in
c903a34347776fe3b89785faa35c446d.
Address some outdated comments and documentation. The property
is read only and calculated, so don't imply that it can be changed
from its default value.
Change-Id: If2dbd9890e533dfccda3eae4cbc96db4f1246f4d
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Test we get the same starts of various days. Some coming changes shall
break some of these, but they should all be back to working by the
time we're taking account of DST before 1970 as well as after.
The first two or three test-cases work by accident in most zones, at
present, due to the zone-based code-path ignoring the LMT period
before the zone's first transition; but Europe/Helsinki had a renaming
transition in 1878, so does see its pre-zone offset between then and
the switch to UTC+2 in 1921, leading to failures in exactly the zone
Coin tests. So suppress these three test-cases pending later fixes.
On Windows, the next text (still pre-epoch) gets bogus zone data for
its LocalTime, so suppress that likewise.
Task-number: QTBUG-80421
Change-Id: I2264e0e436d92112b03264faa410e30057b8f73b
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
qtbug_xxxxx is a fairly ambiguous test name
Change-Id: I4b407160464c9b8300d3683549b0ede837161e7b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
These two tests (applicationTest() and mainWindowTest())
require QApplication::setActiveWindow() to work, which it
does not on Wayland.
Task-number: QTBUG-91418
Change-Id: I0e7b4e24050684b437de63d19bd885bab53d36b9
Reviewed-by: Liang Qi <liang.qi@qt.io>
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
MSVC generates the following for those tests:
warning C4996: '_open': This function or variable may be unsafe. Consider using _sopen_s
instead. To disable deprecation, use _CRT_SECURE_NO_WARNINGS. See online
help for details.
We need to set the define before any C runtime headers are included, so
do it right away, it doesn't do any harm on other compilers.
Change-Id: Ia25afb87934058c3f27e63820eeb2db063a627f1
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
All backend-specific code is now separated and removed
from QSslSocket(Private) code. The original code is mostly
preserved to avoid (as much as possible) regressions (and
to simplify code-review).
Fixes: QTBUG-91173
Task-number: QTBUG-65922
Change-Id: I3ac4ba35d952162c8d6dc62d747cbd62dca0ef78
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
(cherry picked from commit 9391ba55149336c395b866b24dc9b844334d50da)
The _qt_internal_test_expect_fail macro fails when the test fails
during the configuration step. Rename this macro to
_qt_internal_test_expect_build_fail and add
the _qt_internal_test_expect_fail macro that expects inverted test
result.
Change-Id: I4635e99152f7a32f5c48202e84fec59800453d34
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
The tst_QPushButton::sizeHint() test was leaking top level
windows and confusing the other tests, as they may end up
becoming active.
This also re-enables tst_QPushButton::setAccel() since I
cannot reproduce the failure with that, so maybe that was
related to the leaking widgets.
Task-number: QTBUG-91418
Change-Id: I33f4a12bf67f784c5d6a06b2407d067e804d766f
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Liang Qi <liang.qi@qt.io>
There is something wrong with dock widget positioning on Wayland.
In the interest of getting tests running at all, we skip the
test for now, like we already do with a few others.
Task-number: QTBUG-91418
Task-number: QTBUG-91483
Change-Id: I44f5fa0e4d0a43a46ce24a61a899bfcc03787f28
Reviewed-by: Liang Qi <liang.qi@qt.io>
Test that setting QT_USE_PHYSICAL_DPI uses physical
DPI instead of logical DPI.
Change-Id: I7f89cf1af5e013454cc3d8ec3559f2719514fea3
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Add auto-tests for QT_SCALE_FACTOR and QT_SCREEN_SCALE_FACTORS
Change-Id: I57bffa266be910f2ba26cb1a870e09ae202fcae0
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Wayland doesn't support explicit window activation, so the
qWaitForWindowActive() function won't actually wait for the
window to be active.
On platforms like this, we just QTRY_VERIFY() for the window
activation instead.
Task-number: QTBUG-91418
Change-Id: I5f728e6a6bcaac283c026b60e6378b9d5aaf1be5
Reviewed-by: Liang Qi <liang.qi@qt.io>
The tst_qwindow test failed with a warning that programmatically
moving the mouse cursor is not possible with Wayland.
Task-number: QTBUG-91418
Change-Id: I02ceb2af43fbc83a4e6ae09718315f5f79ff8285
Reviewed-by: Liang Qi <liang.qi@qt.io>
The test assumed that a window was registered as the active
one immediately when it was exposed. This created a race
condition on Wayland, where we would often get the wrong
active window when simulating the events.
Task-number: QTBUG-91418
Change-Id: Ie41fe5adb339bbb43d5ee5db400a198e479de386
Reviewed-by: Liang Qi <liang.qi@qt.io>
Several tests failed due to side effects of client side
decorations. We explicitly disable this when initializing
the test to make things go smoother.
In addition, the setToolTip() test tries to set the mouse
cursor position programmatically, which is not possible
on Wayland.
And finally, the qWaitForWindowActive() falls back to
qWaitForWindowExposed() on platforms where explicit window
activation is not supported. This fixes a few issues, but
in cases like focusProxy(), it means we aren't actually
waiting for the WindowActivation event. Instead of testing
for exposed twice on such platforms (Wayland), we replicate
the logic from qWaitForWindowActive() instead and rely on
automatic window activation. If it fails, we do a QSKIP, so
this shouldn't cause any flaky test failures at least.
Task-number: QTBUG-91418
Change-Id: I767c881e7cdc91f43ad357294a2c6240ab1af43c
Reviewed-by: Liang Qi <liang.qi@qt.io>
As QMultiHash uses a pointer for the data, nullptr dereference is a
thing, so check for valid d before doing anything in count()
Fixes: QTBUG-91704
Pick-to: 6.0 6.1
Change-Id: Ia20440cd7bdc03cb09c77f796fb9c5b52765eac5
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This reverts commit 6269438af9, and adds a test.
This change introduced QTBUG-91735, without fixing QTBUG-86857 correctly. The
code already interprets the textAlignment values from the rule, also if no
icon is set. Adding the same, or some default textAlignment to the text flags
if there is no icon doesn't work.
Fixes: QTBUG-91735
Task-number: QTBUG-86857
Pick-to: 6.1 6.0 5.15
Change-Id: Iee07e63a40e72909275f32e1caa28b33a595f879
Reviewed-by: Oliver Eftevaag <oliver.eftevaag@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
The code in qtdeclarative did not do anything at all anymore.
Change-Id: Idd97145cb74aeb4f43dfce2f282a765e90945073
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
There may be a race where e.g. thread 'B' is woken up by a queued invoke.
At the same time thread 'A' asks 'B' to quit, which will set various
atomics (some important ones are 'interrupt' in the dispatcher and
'exit' in the event loop), but it does _not_ try to send another wake
since there is already an unhandled wake triggered by 'B' itself.
Sadly 'B' reads the 'exit' atomic before 'A' updates it.
Then, slightly before, 'B' sets 'interrupt' back to 0, 'A' write 1 to
it, meaning 'A's interrupt is ignored. Then, since there is no
interrupt, 'B' goes back to waiting for events, leaving the thread alive
and running instead of quitting.
Maybe this has unforeseen consequences (one consequence is that it will
return and re-enter the event dispatcher once more, possible
unnecessarily)
Fixes: QTBUG-91539
Pick-to: 6.1 6.0 5.15
Change-Id: Ie6f861f42ffddf4817d5c8af2d764abe9d9103c2
Reviewed-by: Alex Trotsenko <alex1973tr@gmail.com>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
There was a race condition in the test, since the window
might be exposed before the event to activate the
application has been received. And the short cut is not
triggered before the application is active, which
means it could just be discarded and the window would
never be closed.
This happened consistently when testing on Wayland.
Task-number: QTBUG-91418
Change-Id: I40cb143985175f5f2b5405e9502a48475c93074a
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
