Invalidating the filter model may crash while updating
the source model as it emits a layoutChanged().
Just invalidating the filter works as the sorting of items
is not touched.
Workaround this by explicitly using invalidateFilter()
instead of invalidate().
Task-number: QTBUG-103952
Change-Id: I52abda3936e870bb448c420e98a122edecffda7e
Reviewed-by: David Schulz <david.schulz@qt.io>
These type don't support child nodes.
Fixes: QDS-9026
Change-Id: Ic4afd1631db400d8a0efa95480f4a23467fdabf4
Reviewed-by: Mahmoud Badri <mahmoud.badri@qt.io>
- Add "Extension 4" if the name contain "-ext4"
- Fix issue with two Tiramisu packages, now one of them
conatin "Extension 4" in the end.
Change-Id: Ib84807e9401acdef53c3dc1195dc3dc6ef34a57d
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Moved the Effect Maker functionality from assetslibrary model to the
widget, so as to have access to the `directoryCreated()` which is used
on the QML side to (forcibly) update the rows. This should fix it, even
though I could not reproduce the problem for effects myself -- adding
ordinary assets, though, always emits `directoryCreated()` to avoid
this kind of problem.
Task-number: QDS-8916
Change-Id: Ib72c4a235851eed266b6751bb67d4de6acbcfcc5
Reviewed-by: Thomas Hartmann <thomas.hartmann@qt.io>
The previous implementation was saving the _row_ of the TreeView before
performing the rename, and after the rename it was expanding that _row_.
However, when renaming a folder causes it to be moved up or down between
the rows, the saved _row_ number becomes invalid.
The better way is just to use the same mechanism that is being used when
creating a new folder: add the new folder path to a list, and later on
expand rows based on that list.
Task-number: QDS-8896
Change-Id: I6a6bb2faaa7d03ea2af384e5f2b635000577a94c
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@qt.io>
Reviewed-by: Thomas Hartmann <thomas.hartmann@qt.io>
Get rid of QFutureInterface argument from
ITestConfiguration::createOutputReader() and from
TestOutputReader c'tor.
The fine-grained progress reporting was broken anyway:
1. The assumption was that testCaseCount was meant to be
the total number of test functions executed. It didn't
include the initTestCase() and cleanupTestCase(),
while those were reported on runtime apparently
(and exceeding the max progress by 2).
2. In case of tst_qtcprocess, when the whole test was run,
the testCaseCount reported 41, while the real
number of functions was 26 (+2 = 28 for init/cleanup).
3. While the max progress was set to testCaseCount initially,
the corresponding FutureProgress rendered the progress
always in 0-100 range, what didn't match the reality.
Instead, rely on TaskTree progress, which resolution
is per test as a whole. So, when executing a series
of tests this should scale fine. In addition, the
progress advances fluently according to the expected
run time - with 10 seconds hardcoded.
The original code locations, where progress was bumped,
are left with a TODO comment for any possible future tweaks.
Like in case of result reporting, fine-grained progress
reporting may be implemented by providing additional signal,
so there is no need for QFutureInterface inside
TestOutputReader.
Change-Id: Idc11d55e3a49dac8d1788948b9a82f68199203c6
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Enhancing the StyleHelper by some functionality used
on the results pane of the AutoTest plugin and make use of it
inside the output formatter.
This highly improves readability depending on the current theme
when having output formatters and allows to easily re-use the
output formatter there instead of having several stuff
re-implemented.
Mainly relevant for output formatted with ANSI escape codes.
Change-Id: Ic2f5eff877656eb52e3bd2fda0ec9a015e54ea82
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Updating the New project-wizard for qmlproject based projects to use
the stable API as released in Qt for MCUs 2.4.
Task-number: QTCREATORBUG-27721
Change-Id: Ie97d6478135022c1733927c98a3c7f729dc0bb52
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Amends a3153c535d
QT_QMAKE_EXECUTABLE was not taken into consideration and this would
result in not matching kits.
Also Visual Studio generator will not have a CMAKE_C|XX_COMPILER values
in the cache.
On Windows there could be paths with "C:" vs "c:" and the hashes again
wouldn't match.
Change-Id: I5da5fafbd29f3d1fd8c9615c41b7659e63c30ff1
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
As of a3af941adf, the built-in highlighter
can properly handle multi-line raw string literals, so we don't need to
abuse the semantic highlighter for this anymore.
Fixes: QTCREATORBUG-26693
Fixes: QTCREATORBUG-28284
Change-Id: If644767dfa8a97294e84a541eea44143e8d1bb88
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: David Schulz <david.schulz@qt.io>
Get rid of unneeded local variable.
Use boolean OR instead of bitwise OR (regressed during refactoring).
Assert on parentContainer pointer inside task's done handler.
Amends 5dc4cd837f
Change-Id: I2517b80ae894317c52d46326681021d1133fbae9
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Marcus Tillmanns <marcus.tillmanns@qt.io>
We need to generate two sets of linkspecs in order to correctly mark
paths as links in the compile output _and_ the details text of the task
entry.
Change-Id: Ibb4ce89b8104f5edb1f1e45afa27cbd6c3ef5683
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
More inline with the other text block user data members and cleaner on
the client side. Additionally add some & for const arguments and remove
test function call.
Change-Id: I19e646aa204eedf447c02a2f76b89c3430280169
Reviewed-by: hjk <hjk@qt.io>
Adds the option Environment::combineWithDeviceEnvironment. This allows
the linuxdevice to set it to false.
Without this option a deadlock would occur when the QtcProcess tries to
get the device environment and the device shell failed to start.
(e.g. because a required command is missing on the device).
Change-Id: I8f7550d0d90e989a339f27e993a836e67e910e33
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: hjk <hjk@qt.io>
This patch update tooltip text for Button Control fields
in Property editor.
Fixes: QDS-8923
Change-Id: Iac208f56ce3dea14a962abf3c6715de0104d1221
Reviewed-by: Mats Honkamaa <mats.honkamaa@qt.io>
Reviewed-by: Thomas Hartmann <thomas.hartmann@qt.io>
Depending on the presence of the llvm that had been used when building
QC we need to expect its clang.
Use special environment variable to be able to do this correctly.
Change-Id: I2497bbd96483f37aa90842d7fffa5d0a8103ff45
Reviewed-by: Robert Löhning <robert.loehning@qt.io>
Reported by Coverity.
Looks like it cannot really happen in real life though.
Change-Id: I92e09dbe7dc81e694e5450c598c649c956eb7fcc
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
