The current use of AbstractMacroExpander requires some boiler plate
code that can be avoided.
Change-Id: I753d872d824274d94d5105232a8fc05ac7bf0711
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
And reuse that for all .pri files. This is a lot faster.
Change-Id: Iff81a50b892ad666ed6c899aa242b6e9b394f339
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
This unfortunately means doing the filtering by hand,
because filtering on the QHelpIndexModel would be shared
between multiple views.
Change-Id: Iae38952a92dbb1b4a9685aea6f057d96f0d0784f
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
The QSignalSpy parts are 5.4 only, but we need to run with 5.3
Change-Id: I77f778cd66af2f2182d067ca16d2aceeacb9ab3f
Reviewed-by: hjk <hjk121@nokiamail.com>
This reverts commit 1541fbd2ba
in parts. The title bars can now always be shown (the same
state as the old "Unlocked" state) or not always be shown
(the previously "new" behavior).
The default is now "Always shown". Hopefully that makes hiding
the titlebars a conscious (usually one-time) activity
that's not forgotten.
Change-Id: Id31fb193b2eb6376000c845d4a61194e2ed33129
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@digia.com>
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
I'm not 100% sure that this fixes the crash from QTCREATORBUG-12876.
But it's clearly wrong to only check for removed nodes in the project
tree widget and thus only if a tree widget is shown.
Task-number: QTCREATORBUG-12876
Change-Id: Iaaafcef0fe408e8bc2f4d2b61133f4419a612d57
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
Fixes 7 of 9 variables which didn't work due to kit always being null.
Change-Id: I47fbb9ba8aea4736c137939b7ec1d685f94d3fca
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
So that in the context of the buildconfiguration and its buildsteps
current project and others refers to the project of the bc.
Task-number: QTCREATORBUG-12869
Change-Id: Idad7741301432a5fddffdff4225762f4100a3dee
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
There are multiple situations were the snapshot is copied and
access/modifications to the dependency table data is made in other
threads, e.g.:
* FindUsages called by ProcessFile
* findReferences() in cppuseselectionsupdater.cpp
* CppModelManager::projectPartFromDependencies() (clang path)
Change-Id: Ided1c5350f339c5bc960b87b32c78ccc21fa61f6
Reviewed-by: hjk <hjk121@nokiamail.com>
This class is now unused.
Clean up includes of basefilefactory.h (and surrounding includes)
while at it.
Change-Id: I7d4f695b5ebf9ef0a32d523442084d0591f12d02
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
...when adding an include for an undefined identifier.
Change-Id: Ia338e924901262a847d3bd7ed9733d8e66c631dd
Task-number: QTCREATORBUG-10391
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
...in AddIncludeForUndefinedIdentifier. Even without this locator filter
we can offer an refactoring action. See findIncludeForQtClass().
Change-Id: Ic616b82b0af9d59984da882e08f6ede714e85a86
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
Now there is no need to position the cursor on "C" of "Namespace::C" to
trigger the quick fix. It will already trigger if the cursor is
somewhere within the name.
Change-Id: Ic99325f95b5cf5551fc74b55606c402461849fb0
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
...for AddIncludeForUndefinedIdentifier.
There are two basic tests for AddIncludeForUndefinedIdentifier:
* Tests bypassing match() in order to only test the inserting of an
include at the right position (now named *_inserting_*)
* Tests also testing match()
Change-Id: I5b2b008ff9462a1cdaa5718f6a430c46992effab
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
...to make the high level code path easier to follow.
Change-Id: If0ff9a15c9de399286e3a7f59a571949ca8f04de
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
And do the required adaptation in QbsProjectManager.
Change-Id: Ib383ebf99ceb638a545158d426769bf963086a1a
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
... to the editor factories and pass it to the document, not the widget.
Saves some code, puts fewer objects into the object pool.
Change-Id: Iaaf250af74dc4e0c62700873accbb40ba88b7d9e
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
Reviewed-by: David Schulz <david.schulz@digia.com>
There are too many setup without support for --once.
Change-Id: I172b8b8d58a3960d9d93b48e692e612b7627ba56
Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Reviewed-by: Tim Sander <tim@krieglstein.org>
For the time being, this function actually sets an additional
CppIndexSupport to be executed. Handle the case that a client tries to
restore the previous CppIndexingSupport (see SymbolSearcherTestCase).
This fixes GenericProjectManager::test_simple in interaction with
CppTools::test_builtinsymbolsearcher. Now first running the CppTools
tests and then the GenericProjectManager tests work fine.
Change-Id: Ibe7248ec8651d4fdc312f10cdcc9c56d6dd086d9
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
