Instead of re-parsing all test frameworks trigger a reparse
just for the newly added as results for the other should not
have changed.
Change-Id: I6eb53da04bb927a884d6431e137bfe68f391f2ad
Reviewed-by: David Schulz <david.schulz@qt.io>
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
This kind of filtering has been removed long ago
and therefore it should be impossible to get triggered.
Change-Id: Ic9ea02b5446edaadda4d7f21846b37a2f8178505
Reviewed-by: David Schulz <david.schulz@qt.io>
Avoid expanding parsing over all parsers if we trigger a
re-parse for a different parser and there is already a
re-parse postponed.
Change-Id: If74480fea2c671b32083fb7cf3f4dc4c418e6e33
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Reviewed-by: David Schulz <david.schulz@qt.io>
Keep the timer for syncing frameworks where it
is really used, do not create functions that are
not necessary at all.
Change-Id: Ie0a49a2611dc67510b72562455ea2a1af7b1c538
Reviewed-by: David Schulz <david.schulz@qt.io>
And provide an accessor for the framework's priority which
will be needed later on.
Change-Id: I5a2480699d9467ee50e2cc44b9d4ff58f9e8da88
Reviewed-by: David Schulz <david.schulz@qt.io>
More in line with QFileInfo terminonlogy which appears to be
best-of-breed within Qt.
Change-Id: I1d051ff1c8363ebd4ee56376451df45216c4c9ab
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
The scan filter capability had been added to avoid long
scans in bigger projects.
Nowadays the scans are done multi-threaded and the
filtering is probably barely used.
Change-Id: I7d99f5a57bb10deb3d79510db0c7a06bd771c271
Reviewed-by: André Hartmann <aha_1980@gmx.de>
Reviewed-by: David Schulz <david.schulz@qt.io>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Instead, set the default level of all logs to QtWarningMsg.
The call to setFilterRules overrides the user preferences in qtlogging.ini.
Change-Id: Id5f6cd550d14ff7f45ae04c5d3110e0bafb0f072
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: hjk <hjk@qt.io>
It's the more common pattern nowadays, and cheaper.
Change-Id: If6217b2a820fbfa6f088fd9349225f5f8488f593
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
If the code parsers perform a full scan then do not use
cached information.
Change-Id: Ib9635c9715841ab71b97edfe42c4fe9d9d20c23a
Reviewed-by: David Schulz <david.schulz@qt.io>
Also, add context to connect() expressions where we are or were
capturing "this".
Change-Id: I6e006ba6f83d532478018550d148ee93eca59605
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
Reviewed-by: hjk <hjk@qt.io>
Files containing Quick tests are not necessarily listed
inside the project file.
If such an unlisted file is updated we have to scan for tests
even if it is not part of the project otherwise it would not
get updated inside the test tree.
Task-number: QTCREATORBUG-18692
Change-Id: I4c4c67d50b4fff13138b0ab3c884f5ea48a7b8bf
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
As this functionality had been simplified GTest* classes had been
ignored. Continue the simplification and finally remove now useless
code and file.
Change-Id: I89170cd5f05bb93bf30a05fdbf5370012bc9741a
Reviewed-by: David Schulz <david.schulz@qt.io>
Use a thread pool with reduced max thread count.
Task-number: QTCREATORBUG-18185
Change-Id: I18bd3de82365edaf21d8dcf9c89035da1ac74756
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
If a single parser triggers a reparse check if it is enabled
to avoid senseless work.
Change-Id: Ib28f2737be81021e1bfdd33eb347c64ffb5df020
Reviewed-by: David Schulz <david.schulz@qt.io>
If a postponed update should be done for a single parser only and
another full update for all is triggered the latter one would be
ignored which is wrong.
An update for all parsers should always have higher priority and
replace a possible scheduled update for a single parser.
Change-Id: I5e1e446c7dcb9ddbcaed4606ff87a894235b723e
Reviewed-by: David Schulz <david.schulz@qt.io>
* remove unnecessary QLatin1String where possible
* foreach
* nullptr
* bad naming of static members
Change-Id: If0f7c6839be4d2efbfef6f7f0775d2034775bd90
Reviewed-by: David Schulz <david.schulz@qt.io>
This allows to specify folders to be used as search folders
while scanning for tests.
Current approach allows simple folder names or folder structures
without wildcards.
Examples:
Value What will be (recursively) scanned
tests if the current project has any (not necessarily
a direct) subfolder 'tests' this folder will be
scanned
tests/auto if the current project has any (not necessarily
a direct) subfolder 'tests' and this folder has
a direct subfolder 'auto' the 'auto' folder will
be scanned
If there are more folders which apply to the rules then all of them
will be scanned.
This filtering will not keep the parser inside these folders as it
might be necessary to step into different folders because of
dependencies if the found tests, but the search for entry points to
tests will be limited to these folders.
Task-number: QTCREATORBUG-16705
Change-Id: Ib93465540cd20656d033e16205807aba6830d738
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: David Schulz <david.schulz@qt.io>
This avoids doing internal stuff based on getting respective
signals if there is no active test framework.
The only signal that still will perform some more internal stuff
is triggering a rescan from the context or global menus.
Change-Id: I78f065b6bc77a100676d63248b438d3ebc71809b
Reviewed-by: David Schulz <david.schulz@qt.io>
Remove the 'Always parse' setting and respective special handling.
This simplifies the handling of parsing for tests and removes
strange special handling that was introduced in earlier versions
and led more to confusion on the user side.
Change-Id: Ia4d122ed448244f3cb3876dda9930864afde9c28
Reviewed-by: David Schulz <david.schulz@qt.io>
Instead of purging the current test tree on all full parses, just
purge the whole tree if the user switches the current project.
If an item can no more be found after the full parse it gets
removed anyhow.
This avoids losing the check state on any full-parse.
Change-Id: Ia7bae6ced65a69444adf0d7e2a554d90ddd5caa4
Reviewed-by: David Schulz <david.schulz@qt.io>
Adding API for that to the parallel map/reduce functions
Change-Id: Ic7987d899c124dcb04c1ca110b09fc97576873fe
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Avoid unintentional re-enabling of the code parser.
Handling of the enabled state broke several times before,
therefore separate it from other states of the parser to
avoid breaking it again when not taking enough care while
refactoring or adding features related to states.
Change-Id: If1eb0dd649225f10bfc3bf06f09851649da75983
Reviewed-by: David Schulz <david.schulz@qt.io>
Avoid using global thread pool, use internally provided functions
instead.
Change-Id: Id8d3c72c45d85d4806e3692ef3039c31a0eae7b9
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Initial parsing for tests when opening bigger projects was too
slow. Try to let the parsing process on more than one file at
a time to speed up the whole process.
Project before now
qtbase ~11min ~4min
qtcreator ~2min ~40s
Values above may vary depending on the underlying hardware.
Change-Id: I77eeb189c82c58fce84ffbac67143ec7fa0adcae
Reviewed-by: David Schulz <david.schulz@qt.io>
Sometimes the progress indicator did not stop its animation
when parsing for tests was canceled.
Change-Id: I0e4896f2b391ce6690d079fb8b33972cc5fb07f3
Reviewed-by: David Schulz <david.schulz@qt.io>
If parsing the project is not finished we get an empty list of
files to be scanned.
Just trigger another full scan for such a case.
Change-Id: I84b275f76bc98c8f0ef544fc27d30ebd6f8b43d1
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Avoid fetching all files for the startup project again and again
while code model is indexing files. The indexing emits documentUpdated
signals for all (relevant) processed files.
If code model is indexing we perform a full parse anyhow after the
indexing has finished.
Change-Id: Ibfdb47f403da19d5bab3eb2295447ad451fca2b5
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
If tasks are running while shutting down we might end up in a crash,
so cancel all tasks and handle possible invalid accesses of the
current running processing.
Change-Id: I69f7cac5f44390e322fa301af6d6794270c95c2a
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
qbs files are kind of qml documents which end up inside the
QmlJSModelManager. Ignore them by default.
Change-Id: I7136dee80a5f5bfc1b5262daeb00890df28e3f4b
Reviewed-by: David Schulz <david.schulz@qt.io>
Avoid displaying tests from other projects of the same session
which have been indexed but do not belong to the current
startup project.
Change-Id: I6e0fb62cc97898ef3853ec61580dd2deb5ac64e5
Reviewed-by: David Schulz <david.schulz@qt.io>
Instead fetch it once before starting asynchronous processing and
accept that current state of the WorkingCopy might be not completely
up to date. This avoids a crash that might happen when the code model
tries to update the WorkingCopy while the test code parser fetches
information of the WorkingCopy.
Change-Id: I2a893bc8814090361305657ed3c3d772c7bf07d5
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@qt.io>
Mainly a preparation for enabling/disabling test frame works.
Change-Id: I0bbea7e055a607517ef8193587c229cadf0de027
Reviewed-by: David Schulz <david.schulz@theqtcompany.com>
