So that recalculating the information is not O(n^2) but linear.
Change-Id: I69903e0b5ad321d071804d782ad634a3f300e71a
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
The current item was displayed using the hardcoded "monospace" font, so that the current/system
font and size was not used. We now use start with the same font as the error list, and simply
activate 'fixedPitch'.
Change-Id: Ifcf6f0fc044e47b3a57281f6af2d70b7f1c618e4
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
At the end of QmlProfilerFileReader::load() we have a perfectly valid
set of types and events which we can just pass on to the model as-is.
Change-Id: I6981663f409c4647f4d5ae8a73b5d14cc701017b
Reviewed-by: Kai Koehne <kai.koehne@digia.com>
With EngineControl multiple QQuickViews can be profiled in one
profiler if they're all running in the same thread. As the QML
engines are never run in parallel then this results in useful
data.
Change-Id: I83a34fb81fd466c5cac838b096c1b1f322ded882
Reviewed-by: Kai Koehne <kai.koehne@digia.com>
Not doing this makes the QPlainTextEdit interact with an invalid
layout, potentially causing crashes. Now the cursor setting itself is
not part of the undo stack anymore, but as undoing sets the cursor
itself, that shouldn't be a problem.
Change-Id: Ic6f8ee9781faed655ac20f058b562c098ced20ae
Task-number: QTCREATORBUG-12987
Reviewed-by: hjk <hjk121@nokiamail.com>
Add make std::string usable as "compact" key.
Change-Id: Idbfcf9d299e2dde392025166a20c3d0ab60239a6
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
Remove the '.' as the VariableManager will append text starting with ':'
to that textFix strings.
Change-Id: I490319fac964ea8fd8078dff90ec81c9bdef8d09
Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
Reported as a dead store by the clang static analyzer.
Change-Id: Id5eb72ea4698a17486b8a05d88ee66f3fd89bcfc
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
Use "%{}" to put a literal "%" into the output of the macro expander.
E.g. "%{}{Macro}" will be turned into "%{Macro}"
Change-Id: I592789e5cd8f2d52df424db679baf7ba04723202
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
The dead store is not harmfull, but add noise when checking for dead
stores that are due to logic errors.
Change-Id: I244c57e3227d075cb625252e8246c22768bc275c
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
When the QTC_CHECK would fail, a nullptr deref would happen directly
after the check.
The dead stores are not harmfull, but add noise when checking for dead
stores that are due to logic errors.
Change-Id: I83db386e3f0de583469e1576e28955c25930b3b1
Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
Such build system files must be filtered out, as they do not appear on
the right-hand side of "files" properties and have completely different
remove semantics.
Change-Id: I3963aa853003f4d674392434529dab19749af25b
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
createConfigWidget() calls KitManagerConfigWidget() calls discard(),
which results in a signal unmanagedKitUpdated. In response to that
signal, the KitModel iterates over all kit nodes. The node we are
currently creating doesn't have a widget yet, and thus leads to a crash.
Prevent that by adding the node only after the widget has been fully set
up.
Change-Id: I5b14be1c7363e480e3dd938853ff06c595993fc2
Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
The dead store is not harmfull, but add noise when checking for dead
stores that are due to logic errors.
Change-Id: I18084c357696311b6556c7bc60832a72b3318e92
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Since we're opening the port for IPv4, we must also use an IPv4 host
address.
Change-Id: I73e009ebd6deb0bd6804cddc8c83299dbfc1b312
Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
The dead stores are not harmfull, but add noise when checking for dead
stores that are due to logic errors.
Change-Id: I8bf61c51207f0339911bc07fcfbc06f8066521e5
Reviewed-by: hjk <hjk121@nokiamail.com>
The divide by zero is intentional, so suppress the not-quite-so-false
positive from appearing.
Change-Id: Iba96682e5b584a7e4a8e5895af0ccfc68b73cc71
Reviewed-by: hjk <hjk121@nokiamail.com>
This reverts commit c228b36ccf.
There are multiple issues with the reverted patch:
* Infinite loop due to indirect recursion e.g. when opening projectmodels.cpp
* Crash when executing CppTools tests on Windows (no infinite loop)
Change-Id: I38f02132ca57d3d32085db6146d0df7d620d7618
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
The ast parameter cannot be null, because the only use checks for it not
to be null. However, if it would ever be re-used somewhere else, the
logic is plain wrong. Clarification by an assert makes it clear what the
intent is.
Pointed out by the clang static analyzer.
Change-Id: I2c8cba5e5847fc1f92c10021109c55ff8ccd58c4
Reviewed-by: Fawzi Mohamed <fawzi.mohamed@digia.com>
This was overlooked in the original patch.
Change-Id: I1709e28552ed9dd83bd3e810ba51ef5ec95d8580
Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Channel exit is completely normal if we closed the channel, so this must
not cause an error message.
Change-Id: I57f5165b339a52a25118ad5e357f41334dadebc4
Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
“Move Component to Separate File” and “Wrap Component in Loader” did
work only for UiObjectDefinitions, extended them to UiObjectBindings.
Task-number: QTCREATORBUG-12904
Change-Id: I5216110c1edfc6e4536f83eba39e74919c918d50
Reviewed-by: Tim Jenssen <tim.jenssen@digia.com>
The statement in the RFC was misunderstood.
Task-number: QTCREATORBUG-12884
Change-Id: Icd533362ebf95d963cacc8d11da03360e742f070
Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Especially when this gets more complicated with additional fancy windows
mime types and custom mime type for opening a file at a specific
location (dropping from Outline, Type Hierarchy et al), we should not
create that complex mime data everywhere by hand.
Change-Id: I37f6ceb287b0cd055501fdd033ac29816434a020
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
When the check would fail, it would result in a nullptr deref 3 lines
down. As this is a constructor, doing a QTC_ASSERT(..., return) would
leave the object partially initialized, with some fields containing
random garbage.
Change-Id: I823d52d43d90eaf06cb7d7cd64fbee17e5fc889e
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
