Move towards using aspects and away from IOptionPage's QObject
base.
Change-Id: I07850a6b8e9fa7d2591efc6f76a23f249dfc164f
Reviewed-by: Orgad Shaneh <orgads@gmail.com>
Amends 2f39b51bdc
The default tool's value is taken as default global value,
then will be saved for all tools.
This fixes the case when the false Autorun value for the default
CMake tool would always be set as global autorun with no option
to actually set global Autorun value due to the "upgrade" path
mechanism.
Change-Id: I17076bc0c77b087c5d4048fdfe74ddf91d837fd4
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
The base will be cut soonish, and the only use here was tr() with
a (wrong) context. Use StudioSettingsPage for that now.
Change-Id: I41e4b737713591e002c126bbf97226de7c3f784f
Reviewed-by: Mahmoud Badri <mahmoud.badri@qt.io>
The indirection is not needed. It's not really stand-alone, and not
touched before GitPlugin::initialize().
Change-Id: If8552dc69ff9efa18a3f1d6d998a8f28ab65e501
Reviewed-by: Orgad Shaneh <orgads@gmail.com>
The interfaces make much more sense now, and the coding style is taken
into account (except for the fallback case where the user enters the
type).
Change-Id: If08dfc41ebd63287fb5b0b187944e7fdf08b0823
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Amends d8be2491a5
The change above introduced FileApiReader::takeCMakeFileInfos and uses
it to move the information about additional CMake files out of the
FileApiReader. But that now emptied variable was later used to inform
the project about these additional files. So, that broke the automatic
running of CMake when project files (except the toplevel one) changes.
Instead use the list of additional files that now lives in the
CMakeBuildSystem for that purpose.
Change-Id: I1062593029880af9d4c70e72e1bd101d40ad0c00
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
By themselves not worth a pimpl, but since we have one already...
Change-Id: I867e68c5e82f71ebbe16a55eb190d87f981a550b
Reviewed-by: Orgad Shaneh <orgads@gmail.com>
The connect doesn't work with incomplete type FakeVimHandler::Private
anymore, so move the code.
Change-Id: I686cd19a985f965cebf7d0a927cff4dc80ae746d
Reviewed-by: hjk <hjk@qt.io>
(cherry picked from commit e56e3b6f37)
Babystep towards removing QObject inheritance of IOptionsPage
Change-Id: Ib36b8ce4dd1f6878abdfe20c0c4e3ddb95e927dd
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
... in document processor callback. The dependency table seems to get
rebuilt more often than expected, making this operation expensive.
Amends 6d805195f6.
Change-Id: I1eace180b2428756b648fa1c58eb0fb66cce3c0c
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
Operating on saved settings is less error prone and easier to maintain.
Change-Id: I92e3a6b52296cddc302ba2e4410edfd243b8ad32
Reviewed-by: hjk <hjk@qt.io>
Delay the update: The implementation in Qt is currently very slow, and
while we don't have a fix, editing with immediate update is very
sluggish.
Also do not update the preview, if it isn't visible.
Change-Id: I35e10af232064b02a36a8be3758ea0bbc5b66f40
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: David Schulz <david.schulz@qt.io>
Do not watch directories unconditionally, but only if they contain
removed files to check whether those files are readded. This should
reduce the number of needlesly watched directories to a minimum and fix
performance regressions introduced by
61598eca15.
Fixes: QTCREATORBUG-28957
Change-Id: I8fe387e7de32b0fb585074330c7f6ca7eae44730
Reviewed-by: <github-actions-qt-creator@cristianadam.eu>
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: hjk <hjk@qt.io>
... to GTestSettings to prepare for removing the QObject base
of IOptionsPage.
Change-Id: I2d6f78e4f2a1d3690860c40b4ab55008034457a9
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Not needed currently but will be once an IOptionPage loses its
QObject base dependency.
Change-Id: Ide7de1a34ff62c32d63a77040e9ec59bb42e3c40
Reviewed-by: David Schulz <david.schulz@qt.io>
The connect doesn't work with incomplete type FakeVimHandler::Private
anymore, so move the code.
Change-Id: I686cd19a985f965cebf7d0a927cff4dc80ae746d
Reviewed-by: hjk <hjk@qt.io>
fetchMore is called multiple times before expandNode and seems to
invalidate the index that is passed to expandNode. Since we only need to
fetch more when we want to expand the item we might as well just
integrate the code of fetch more into expandNode.
Fixes: QTCREATORBUG-29083
Change-Id: I0e60e9bb03b53de2e86eea232fb5bb98046bbb80
Reviewed-by: hjk <hjk@qt.io>
Current Project is misleading. Current Repository refers to the
repository of the current file, which is much more expected for
"everything around" actions (Diff, Log).
Task-number: QTCREATORBUG-10170
Change-Id: I252f40cd3dfd9de184a889355b59f91bb64983dd
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
QTC_SCOPED_TIMER() is a handly debug tool that measures
the time spent in a block of code. It starts measuring the time
when the line where it was placed is reached, and stops measuring
the time when the current invocation exits the block in which it
was placed (i.e. when QTC_SCOPED_TIMER() goes out of scope).
The QTC_SCOPED_TIMER does two printouts:
1. When it starts - it prints the current time, filename and
line number.
2. When it ends - it prints the current time, filename,
line number and the timeout in ms.
The QTC_SCOPED_TIMER() was added into qtcassert.h file on
purpose, as this file is included already in most of the
codebase. In this way, when it needs to be used, it's enough
to add a QTC_SCOPED_TIMER() without adding extra #include.
Example use case, after adding the "QTC_SCOPED_TIMER()" into
ProjectExplorerPlugin::initialize() as a first line:
SCOPED TIMER [14:46:57.959] in [_long_path_here_]/
projectexplorer.cpp:823 started
SCOPED TIMER [14:46:58.087] in [_long_path_here_]/
projectexplorer.cpp:823 stopped with timeout: 127ms
Change-Id: Iaed3f297c8aeb6e90dd9909e76fc9933599a39b6
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Intenally just wrapping a 'bindTo' result, but less trigger potential
for pointer related peladophobia
Change-Id: I25171a2675fb0474ce97c04552ac1cf5ffd6ee56
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Reviewed-by: hjk <hjk@qt.io>
For better performance, include also the scheme and host in qHash
calculation.
Change-Id: I2a69a128597429b88a71943d248ce038b49285f2
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: hjk <hjk@qt.io>
Currently almost the same as the comparison example, but planeed to
grow.
Change-Id: Id129266a1b5b12438e2fb643ee114e6d816581f6
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
