Qt 6 brings a behavioral change regarding delivery of wheel events:
92df790f46b3a8b17aec2f385d6472fd3f8647f6 and
2a857ee28315c5bacfe2ecaf402ca9005b35c20e. If there is an event
filter installed and wheel event is being processed, the event
is now accepted by default, which means that after returning
true from eventFilter(), the event is not propagated anymore.
Since we want to redirect the event to the parent HelpViewer object,
we explicitly ignore the event prior to returning true.
This patch fixes the QTextBrowser backend.
Amends 0b96a1b76b
Fixes: QTCREATORBUG-26369
Change-Id: I7a0167c54f830fffab13d73eb91cb2ecbd730c86
Reviewed-by: Robert Löhning <robert.loehning@qt.io>
- We checked for the QFuture's canceled status at a time where it would
always yield true.
- A pointless default value was used for the project root path.
Fixes: QTCREATORBUG-26356
Change-Id: I4d17c6ecb7e9e7e71df737d85b42feb37d3ead5d
Reviewed-by: David Schulz <david.schulz@qt.io>
Use the specialized operator== of ToolChain when looking for priorly
defined autodetected tool chains.
Fixes: QTCREATORBUG-25697
Change-Id: I477d29695eef2acb1ba566d42df9b1e57d26eb07
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
... with regards to the session include paths.
Amends 0636238429.
Fixes: QTCREATORBUG-26323
(cherry picked from commit 8c86b9bca1)
Change-Id: I972a314db607d0b06859744dec4c83e885be3b5c
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: David Schulz <david.schulz@qt.io>
...from doc/qtcreator/images to doc/qtdesignstudio/images
Change-Id: Iaa49fbb4d620a833edd41b97874aad631b7c25b5
Reviewed-by: Johanna Vanhatapio <johanna.vanhatapio@qt.io>
LLVM moved to github a while ago - the old svn repositories are dead.
Furthermore, (almost) all of the code is now under the "Apache 2.0
with LLVM Exception" main license
(see also https://foundation.llvm.org/docs/relicensing/)
Change-Id: I8319f8479357c6182c5bf78b1ee7fe202e666980
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
These were never used. And updated ones are upcoming with the new QDS
application wizards.
Change-Id: I998c3c0b8cfa3033d8be46524db3ed978738e729
Reviewed-by: Artem Sokolovskii <artem.sokolovskii@qt.io>
Reviewed-by: Samuel Ghinet <samuel.ghinet@qt.io>
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Usually that is done by Qt, but not if reduce_relocations is off.
Fixes: QTCREATORBUG-26435
Change-Id: Idd337d59555fd6e732215b9c753820672f63c91f
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Christophe Giboudeaux <christophe@krop.fr>
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
In some unlikely circumstances it may happen that
HighlightingResultReporter might be destroyed before
returning from the HighlightingResultReporter::start()
function. This may lead to undefined behavior.
Refactor the HighlightingResultReporter so that instead
of using QRunnable subclass we define simple function
returning the QFuture object directly.
Change-Id: Ib833771a7e46e87c83d10b59ca056a0147fabe88
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
The logic for applying formatting was completely broken and had
obviously never been tested with a formatter that does anything.
Fixes: QTCREATORBUG-26420
Change-Id: Iffc1784b225969b86b221afb16c06e8e0d053b91
Reviewed-by: David Schulz <david.schulz@qt.io>
On Windows QtCreatorAPIInternal.cmake would contain a check for
BUILD_WITH_PCH and if not set add a few defines to the DEFAULT_DEFINES.
First time the BUILD_WITH_PCH would not be set, because the option
will be set later, the DEFAULT_DEFINES will contain the specific non-pch
defines.
The second time BUILD_WITH_PCH would be ON and those non-pch
specific DEFAULT_DEFINES will be removed, and the build system has
to do a full rebuild because the defines changed.
Change-Id: I3f039a91667affc35f18103cfed062481f9dc93e
Reviewed-by: Marco Bubke <marco.bubke@qt.io>
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
The arguments of the synchronization function are changing constantly, so
all tests have to adapted even for default arguments. To reduce this
changes now a struct is used as parameter.
Task-number: QDS-5196
Change-Id: Id3b6a530bcb8ab043eccce6526088f345d746235
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
There are now different modules for qml, cpp and directory imports.
Cpp imports are qml import append by "-cppnative".
Task-number: QDS-5196
Change-Id: I599b224892c5ada0224a765d0bde8b32de20b061
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
The type will be use source id and name instead of module id and name as
key.
Task-number: QDS-5236
Task-number: QDS-5238
Change-Id: Ibc9c298dc0a6363b630173ec4981d574cecd02ff
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
References to these files have been removed, but the files
were forgotten.
Change-Id: I412d5daf2acf07b4f8c00715bc01e4422bbc1179
Reviewed-by: Johanna Vanhatapio <johanna.vanhatapio@qt.io>
This fixes the recursive expansion of e.g.
"qtbase/tests/auto/corelib/io/qfile/copy-fallback.qrc"
which includes itself.
Change-Id: Ib345a2d6de4fa2eefd76eed58b99a72fcb8c84b5
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Detect that constructor of ResourceTopLevelNode is being
run from non-main thread and omit creation of ResourceFileWatcher
in this case. The construction of ResourceFileWatcher
will be postponed until the node tree returns back
to the main thread. This happens later inside
Project::setRootProjectNode() when ProjectTree::applyTreeManager()
is called for the second time - this time it's done from the main
thread. In order to setup the lacking resource file watchers
we install an additional folder node visitor only in case when
the handler is called from main thread. The visitor
sets up the lacking resource watchers if that's still needed.
Amends: 0bcab32657
Fixes: QTCREATORBUG-26417
Change-Id: Ia1bfb7f284afb833b6b4291accc4d0a91bd0d6c5
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
... for clangd.
Putting it in the build directory seems sensible in principle, but that
can be problematic for in-source builds. So introduce another level of
nesting to prevent conflicts.
Fixes: QTCREATORBUG-26431
Change-Id: Id66aa0852d206695f2fc2ec42292b1cecefe2b59
Reviewed-by: David Schulz <david.schulz@qt.io>
Locator tries to de-duplicate entries in case results from multiple
filters are collected. This is done so we don't get e.g. multiple
results for file names, from the various file filters (open documents,
project files, included files, ...).
For this we define operator== and qHash for LocatorFilterEntry, and add
them to a set of "seen" items based on their "internalData" QVariant.
For string data that works fine. For non-string data we used qHash for
the variant's internal data, and that changed in Qt 6 so we get the same
hash a lot of times which hits on the performance big time.
Defining qHash generically for QVariant isn't really supported. Since
our use case is to de-duplicate items if the internal data is a string,
simply restrict our de-duplication to that explicitly.
Fixes: QTCREATORBUG-26415
Task-number: QTCREATORBUG-24098
Change-Id: Ifdad17e7e66e9fe4b1300de51c45a5efac73cf3a
Reviewed-by: David Schulz <david.schulz@qt.io>
Moved showing and hiding categories logic to Cpp side so that:
- Currently selected category is persisted when the component library
model is reset.
- QML side is clean of the mix of logic and UI.
Also reworked some logic/variables that are making the logic complex.
Task-number: QDS-5215
Change-Id: I8e9f5893f37a982283f1b1be9fee022f0b8afa32
Reviewed-by: Samuel Ghinet <samuel.ghinet@qt.io>
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@qt.io>
When asset importing fails we show a warning, but not when the user
cancels the importing.
Change-Id: I950d3b43c0f7c78a5abfb31a1974d2a9cd193522
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@qt.io>
"erase" without namespace conflicts with std::erase, with the latter
taking precedence.
Amends 9929d3dd73
Fixes: QTCREATORBUG-26386
Change-Id: I7fa64827ad61f1da262ce48082854975bc431c69
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
Since we employ QtcProcess, the reaping of possibly
running process is done internally by QtcProcess already.
Change-Id: Ie65314a4aecf5bb1808b0f2c46a50503d82fd6fb
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
It's often interesting to know how long it takes for a reply to a
request to come in. While that data is in principle already available
from the inspector, a dedicated logging category gives much quicker
results.
Change-Id: Idd43c6e69422140863a7adaa33a7bc17dd8e6492
Reviewed-by: David Schulz <david.schulz@qt.io>
... with regards to the session include paths.
Amends 0636238429.
Fixes: QTCREATORBUG-26323
Change-Id: I8e2cd5f5e87d9dc3d2df3f943e13599bc7139768
Reviewed-by: David Schulz <david.schulz@qt.io>
... when highlighting with clangd. This can happen due to macro
invocations or unusually placed includes.
Task-number: QTCREATORBUG-26396
Change-Id: I471faee13c62fa511bdedfdd5b864327e858f6b8
Reviewed-by: David Schulz <david.schulz@qt.io>
