| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
Task-number: AUTOSUITE-1515
Change-Id: Ia2df5f7c94dfdf8aef9f82eecd002fcb938fbca1
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
| |
Change-Id: I6ae1a8a7f3f2f1f2ce74aaf1b60401fd34a6da77
Fixes: AUTOSUITE-1520
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make sure to also use the fromRemoteObjectVariant() helper function
when emitting the current state directly from the initialize() call.
Otherwise the QIviMediaPlayer will bail because it can't accept the
broken matroska QVariant.
Change-Id: I162c90cb64e6a81a7c85674783fb379deea910df
Fixes: AUTOSUITE-1475
Reviewed-by: Egor Nemtsev <enemtsev@luxoft.com>
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
With virtualenv 20, the --relocatable flag got removed
and causes build failures in QtIvi. As we still need
this functionality for making our virtualenv work
after 'make install', we copied the legacy code and
converted it into a standalone python script which
does the job for now.
With virtualenv 20 the complete deploy mechanism changed
and is now based on top on venv. To make the new environment
relocatable we need to copy additional files as well as
read the pyvenv.cfg instead of the old-prefix.txt
Change-Id: Ib77bbd2a1c959693bf1ae6ca478086ce2c848036
Fixes: AUTOSUITE-1482
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
(cherry picked from commit f2be418701f451a4d68a4949d996b873b1836305)
|
|
|
|
|
|
|
|
|
|
|
|
| |
This makes sure pinned versions are used when installing all python
dependencies, which is needed as a recent jinja update causes problems.
Change-Id: Ia6cfed8169aca8eeac6809807772a448199c433c
Fixes: AUTOSUITE-1440
Reviewed-by: Vladimir Minenko <vladimir.minenko@pelagicore.com>
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
Reviewed-by: Aapo Keskimolo <aapo.keskimolo@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
|
|
|
|
|
|
|
|
|
|
|
| |
Usually we always use a commit in the upstream/develop branch,
but for the 1.10 bugfix, we need to switch the branch.
To make our life easier, remove the tracking branch completely
to accept any sha in the CI.
Change-Id: Ieb9e003691d4ffab0e1f73737b70936574066877
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
Because we need a QGuiApplication for playback on macos, we also
get an app icon in the dock.
Use QT_MAC_DISABLE_FOREGROUND_APPLICATION_TRANSFORM to prevent this.
Change-Id: I7806e4c8e113c22820cddafa19db34b8bec57a50
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Using a singleShot timer is not the right way as it doesn't make
sure the call is really executed in the next event loop run.
Instead use QMetaObject::invokeMethod().
Also reorder the qivipendingreply autotest to make sure the QML
test is done first, which needs the types to be registered...
Change-Id: Id4f8dd1577e653c99fd82e8ab8f93a848f9b1b79
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
When QtIviCore is loaded at runtime the init function
is executed right away before all types are initialized.
This causes a crash on windows and we now delay the
initialization until the library is ready.
Change-Id: I979f48c811f6a7f10facb59ef381239d5ea80367
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Only in the DataChanged mode we want to use the onCountChanged signal
from the backend to fill the model with placeholders, as the new
fetched Data will replace these.
This fixes the check to ignore the signal as long as the FetchMore mode
is used.
Change-Id: I323f6f1c04d7386a5ca35d1bdd3ab2cff13bacdd
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
| |
The constraint_string() function doesn't exist, instead we need to use
constraint().
Change-Id: I299e6bd7e71a32cb48c0c8d69ad26f054d4094a3
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
| |
This is needed as otherwise a QPendingReply is not accepted as a
return value of a function called from QML.
Change-Id: Iaaed897aa0c49874e705779032d4cd14d4664998
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
| |
Change-Id: I50b0b4bbe98272c2a6d617f7b7c62cb906d6fe81
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
| |
This partially cherry-picks:
6e329be7d833279f3e616fb975de89332d3b97c1
Change-Id: I84e329d27828c740a57cf95545a0a1e45cd0ad6a
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
This also adds a meaningful error message in case the class providing
the enum is not registered to the meta type system.
Change-Id: Id0f64df44b37926ee55c33b4663498512f8923df
Fixes: AUTOSUITE-1403
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
* Added a link to the Generator Tutorial to the front page
* Added a short paragraph about RPATH linking to the tutorial
* Fixed linking warnings
Change-Id: I085ddbe5c0e9939e2e22ae7188c307fae52f77b2
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
| |
From inside the QML simulation it was not possible to properly iterate
over all items of a model.
The missing count property is added by this commit. In addition the
at() method is fixed to return the Q_GADGET based type by copy instead
of returning a const ref, as QML cannot cope with that.
Change-Id: Id12f05814684d3e9d482601fbaa788dc9c450df3
Fixes: AUTOSUITE-1371
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|\
| |
| |
| | |
Change-Id: I9a2ff22058e273be9334d9ca5a4b86ece7d9626b
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Due to a change in QtCore, it is not possible to call
QMetaObject::newInstance() on Q_GADGETS anymore.
Because of this, creating qface based structs from
simulationData resulted in the error:
QMetaObject::newInstance: type Contact does not inherit QObject
This commit changes the way how the structs are created
and is using QMetaType::create() instead and instead
of using a constructor taking a QVariant, it invokes a private
fromJSON() function, which now gets generated instead of
the QVariant constructor.
Fixes: AUTOSUITE-1374
Change-Id: Ibef59fcc0a209de3d98f055eaed4e4c04521575f
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|\ \ |
|
| |\|
| | |
| | |
| | | |
Change-Id: I576fbcc52a2f7eb50c6303d2296f9daba16618fb
|
| | |
| | |
| | |
| | |
| | |
| | | |
Fixes: AUTOSUITE-1357
Change-Id: I52e7affea70bda6416be4267149a33a6393e1627
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Access to static object pointer without checking.
Object might be already destroyed.
Change-Id: Ida459bc0e41f0b475293f2b69c44789f535df549
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
In chapter3 we included the wrong snippet for the main.cpp.
Change-Id: Ia9f89e5b2ea14c82b70aee5937fd9320b980ed7f
Fixes: AUTOSUITE-1351
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
For default_values in our simulation backends, QLatin1String() were
generated so far, but when using these for initializing a struct
with only one string property, this caused problems, as the structs
are using QStrings and because of an additional QVariant constructor
this caused some ambiguity for the compiler.
Switching to QStringLiterals fixes this problem, as the QString
constructor is used in this case.
Change-Id: Iae34ef290ec5683665fdfdb75b633e748ee371f3
Fixes: AUTOSUITE-1340
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
For default_values in our simulation backends, QLatin1String() were
generated so far, but when using these for initializing a struct
with only one string property, this caused problems, as the structs
are using QStrings and because of an additional QVariant constructor
this caused some ambiguity for the compiler.
Switching to QStringLiterals fixes this problem, as the QString
constructor is used in this case.
Change-Id: Iae34ef290ec5683665fdfdb75b633e748ee371f3
Fixes: AUTOSUITE-1340
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|/ /
| |
| |
| |
| |
| |
| |
| | |
In chapter3 we included the wrong snippet for the main.cpp.
Change-Id: Ia9f89e5b2ea14c82b70aee5937fd9320b980ed7f
Fixes: AUTOSUITE-1351
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
| |
| |
| |
| |
| |
| | |
Fixes: AUTOSUITE-1332
Change-Id: Ie6f5f803a68b2945647973d23c9197b9fc4855e0
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
| |
| |
| |
| |
| | |
Change-Id: I2d65ce049de503d3b22ca15583815f8ff547a4ce
Reviewed-by: Dominik Holland <dominik.holland@qt.io>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This is needed for the plugins to find the frontend lib on the target
Change-Id: I6b83b5ac96b450f7a35b241b2f64275d2ff5dbc5
Fixes: AUTOSUITE-1250
Reviewed-by: Stanislav Maksimov <smaksimov@luxoft.com>
|
| |
| |
| |
| |
| |
| | |
Task-number: AUTOSUITE-1250
Change-Id: I6cbdc49fce071d4b5e5e415d313d3622362e47c8
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
By using Q_EMIT, Q_SIGNALS and so on, then this ensures that the
generated code will work with no_keywords set for a project in a pro
file.
Change-Id: Id53e2219ac215e5b6bef99f835b2ffd42a59ab4e
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Versions of Python before 3.7 will escape any character that might
be used in a regexp when using re.escape() even if it is not
completely part of a regexp expression. For example : will be
escaped as \: which is invalid when used inside a string. Therefore
we should just ensure that any instances of a \ inside a string is
escaped correctly when generating the C++ code by using replace().
Change-Id: I746d99bf6bb7a9b69a1d4747da49ad0fb85f151d
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
files
Generating a file called "main.qrc" has some strange side effects during
compilation. Making sure the file has a postfix circumvents this.
Change-Id: I2a2bce686154e2659df997a0f3e09a6bfbd571d7
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
'getter-name' correctly
The underlying QtRO transport doesn't allow us to define the getter or
setter names for the property, that's why we need to use the property
name when deriving from the autogenerated source classes.
Change-Id: I417e85af5460d51fab9964fc163279e5eea3afe1
Fixes: AUTOSUITE-1242
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Method parameter interface conflicts in MinGW build.
Renamed interface to interfaceName.
Task-number: AUTOSUITE-1236
Change-Id: I18a5841810d8ce784872cecd41469e34a315a034
Reviewed-by: Teemu Holappa <teemu.holappa@qt.io>
|
| |
| |
| |
| |
| | |
Change-Id: Id663c3d735a3324e2a47a8174c2e5cf8b01f0263
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Method parameter interface conflicts in MinGW build.
Renamed interface to interfaceName.
Task-number: AUTOSUITE-1236
Change-Id: I6e425830590ab8281772ae1f3372fa9a2410626f
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| | |
Change-Id: Ie41f7d15bdcc3c2bdcf7de299063d016186bd3a1
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| | |
Change-Id: Ia0e72321c6252b15149d23cbd9c0ceacc67fd775
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| | |
Change-Id: Id90e9623db6e6c72a89553f2d7c85004526776df
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
The plugins as such don't have a specific version dependency
and the dependency it had before was due to a problem in qmlplugindump
Change-Id: Ia07b6ed4cc75e6bab27554dc8c08373e38ca3418
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
|\ \ |
|
|/| |
| |/
| |
| | |
Change-Id: Id96d0a6d670c2e9918e4feeed4e3da3b0ac2512f
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This reverts commit 83720c2961b7af019643f5318d84a4ea443a5a87.
This change was breaking SC and can't be released with a
patch release. The change will be put into the dev branch
instead.
Task-number: AUTOSUITE-1187
Change-Id: Icfbf6e34dddbb93160681c97f53abef7915aba78
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
During the Qt cleanup, some parts of QtCore are still using qDebug and
friends to e.g. when QT_DEBUG_PLUGINS is set.
Because of this the messageHandler might be called while the
QDltRegistration is already destroyed. In this case the messageHandler
can't log anymore, but it also shouldn't crash because of a nullptr
access.
Change-Id: I4d75acc5dfacca89d4876c13580c9d7bf2838f69
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Introduce a mutex to prevent issues when registering
a new context/application while logging from another
thread.
Change-Id: I734b2a363ffe83fc9e9728898707cdefef689824
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
DLT silently drops too long messages and produces
empty lines instead.
This changes introduces a behavior enum which helps
to configure geniviextras to truncate or split the
long messages.
Fixes: AUTOSUITE-1187
Change-Id: Ieece74a984460fb6c2fe603789361b59ada9df12
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| | |
The function was defined already before, but it didn't have an implementation.
Change-Id: If234d83cbcd15bee6e24459c9bd27ff9d6ae05a2
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| | |
Change-Id: I8dbc90830f815b90f26af7c343b3548c13c60d94
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|