remove(int, int) with O(N) was used in a loop.
We had a quadratic complexity.
Use erase-remove idiom to fix it.
Change-Id: I643a2a75619ec5ea2bf99e48a25f64a7f69ba156
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Add new file for dynamically loading Window 10 Helpers via combase.dll
and add function qt_windowsIsTabletMode() querying
UIViewSettings::get_UserInteractionMode() for "Touch" mode.
The style hint QPlatformIntegration::ShowIsMaximized will then
reflect the tablet mode setting.
Task-number: QTBUG-56831
Change-Id: Ia361dd172fcf0e54fdfc70863c43527f3ea72fe2
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@qt.io>
By templating on the <chrono> types and unconditionally using
duration_cast to coerce the duration into a milliseconds, we
allowed code such as
mutex.try_lock_for(10us)
to compile, which is misleading, since it's actually a zero-
timeout try_lock().
Feedback from the std-discussions mailing list is that the
wait_for functions should wait for _at least_ the duration
given, because that is the natural direction of variance
(tasks becoming ready to run might not get a CPU immediately,
causing delays), while an interface that documents to wait
_no more_ than the given duration is promising something it
cannot fulfill.
Fix by converting the given duration to the smallest number
of milliseconds not less than the original duration. If that
is not representable in an int, use INT_MAX, emulating the
effect of a spurious wakeup, which are allowed to happen if
the function returns false in that case.
In the above example, the try_lock_for call is now equivalent
to
mutex.tryLock(1);
The tryLock() docs state that the actual waiting time does
not exceed the given milliseconds, but fixing that is a
separate issue.
Change-Id: Id4cbbea0ecc6fd2f94bb5aef28a1658be3728e52
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Found by ASan when it should have been found by me in the initial
review...
The old code did, in this order:
1. delete item->layout() (which deletes item, as QLayoutItem::layout()
is just a dynamic_cast implemented with virtual functions)
2. delete item->widget() (which is correct, but too late, as 'item'
may have already been deleted; this is the first use-after-free
bug)
3. delete item->spacerItem() (which is the second use-after-free).
Fix by deleting item->widget() (which may be a no-op), _then_ checking
for a QLayout, deleting all children (but not the layout), and finally
deleting item as a QLayoutItem.
Rename clearQLayoutItem() to clearAndDestroyQLayoutItem() to better
match what it actually does.
Change-Id: Id70a7a137dac5de466ef619f01bfd61dfc150418
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Samuel Gaist <samuel.gaist@edeltech.ch>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
The time stamp is added at the end of the node information and
consequently this also bumps the version.
Task-number: QTBUG-57182
Change-Id: Ia10e006f28c0b168b2bcd74ed8b7098f84d10af3
Reviewed-by: hjk <hjk@qt.io>
Change the main window to contain a QTabWidget and add a log widget
logging relevant events on the top level widgets for testing changes.
In the preview window, add new window flags of Qt 5 and output
geometry, margins and window state in addition.
Change-Id: Icec366223b6c163d58a69034687f3d9323a91533
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Chaining hooks together was mentioned by Ossi in the comments of
d953d9a4. This patch justs add a test that verifies that it works, and
also serves as an informal example for developers looking how to do it.
Change-Id: I53a014d5663c289ea0559e0926ed301f4e5110e6
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: hjk <hjk@qt.io>
This will be used in QtQuick to avoid costly string manipulation (which in turn
involves memory allocations).
Change-Id: I51a67a4cd97cc576f399483c9c0c13da1e1c6e72
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
When a menu item's enabled state changes after
-[QCocoaMenuDelegate menuWillOpen:] is invoked, i.e.,
during or after QMenu::aboutToShow() is emitted, that
state change may not be taken into account. This is
because the automatic menu validation, upon which Qt
relies, is not made aware of any such change.
By calling -[NSMenu update] when syncing the QPA menu
item, we induce Cocoa to invoke -[QCocoaMenuDelegate
validateMenuItem:] and ensure that previously synced
items, whose state may have changed, will be properly
updated. This, however, has a small side effect, namely
that menu-holding items will also go through the automatic
menu enabling path and may appear disabled since, until
now, they were not properly configured. In order to solve
this, we set the action on those items as well, and make
sure that both of QCocoaMenuDelegate's relevant methods,
validateMenuItem: and itemFired:, properly process
menu-holding items.
Menurama manual test updated accordingly.
Change-Id: I62f955538b8be09b8494ea0ce87fca7910148d38
Task-number: QTBUG-56850
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
That loop could never have worked properly.
Identified by analyze compile switch.
Change-Id: I5e987dc9f5c57d3ffbcf054b7b417b506c02e7e1
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
this is complementary to configure's makeSpec library source type.
this should be sufficient to make QMAKE_USE += {egl,opengl,opengl_es2}
actually work, obsoleting the need for opengl.prf and egl.prf (and the
currently dysfunct openvg.prf).
Change-Id: I2f7595ac89afa087ea7f0f25060e8e47e6148be9
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The documentation for the qtmain license is right now pretty hidden.
The library is added by qmake independent of the Qt module that are
linked to, but it seems safe to assume people at least link against
Qt Core.
Change-Id: Id474990edde45feab6727bada2bc6a28946216cd
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Topi Reiniö <topi.reinio@qt.io>
Since virtually all the logic is shared between QNSColorPanelDelegate
and QNSFontPanelDelegate, we extract the added buttons and
layouting logic and move it into its own class. This requires
the two afore mentioned Objective C classes to satisfy the
QNSPanelDelegate protocol.
Change-Id: Ie26e758f5db71920896d930a4f3644b51a1ce3fa
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
now the callbacks don't need to re-export the designated file contents,
which improves the abstraction and removes some boilerplate.
Change-Id: Ifa50313155fc96762025e2610b810ebb71daa373
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
swap the branches for (not) reading from qt.conf, and use a state
variable instead of an 'else' for mutual exclusion. this is somewhat
more self-documenting, and allows for a saner handling of the mkspec
fallbacks (which really should have been in a separate [QMake] section
along with Host* and Sysroot, but changing that now is way too much
hassle downstream).
Change-Id: I80a73294022fd1e8d84fe501b737c4fc7758662f
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
so far, qmake.exe was built in qmake/ and then copied to bin/, with
possible errors in the second step ignored. this made no sense.
this unifies the nmake makefile with the unix one; compare 46e51ce1d.
Change-Id: Ieb9c7cd46f0be0501d17e297808ac1cdad1b3c4a
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
when the file's effects are discarded, the mention of the file should be
as well.
Change-Id: I894b7e2b887dd34d18533b197bfa9d0d84d647e7
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
the normal mode forces the prompt into a pattern which may be
undesirable.
Change-Id: I01689c7a6573415801862348b32bafc6a609ed4a
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
that is, adjust path separators and don't quote them. we already did
that to some degree, but totally inconsistently, so it just didn't work
for any targets with "fancy" file names.
note that we don't bother doing that for recursive targets, as these are
assumed to be identifiers.
Change-Id: Ic75f003b71abc6fed03a4121b903ad5ee8253ed2
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
if the externally provided qmake had a different on-device install tree
layout, things would go wrong. of course, that's a rather unlikely case,
so nobody noticed ...
Change-Id: I59f9976a769ccb6099b7237ef42555f0549615aa
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
there are plenty new functions, let's do this symbolic act.
Change-Id: Iaeb88afa5e33cacd81dc0ea26e380a16af06a739
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
otherwise it's sometimes quite hard to tell it apart from configure's
own messages.
Change-Id: I2f4908344367a9a3ce38e032bf76486fc4552ffd
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
this got broken in 2ad4d75754.
however, the new configure system operates from the top-level build dir
anyway, so there is no point in messing with the cache as a reference
point to start with - just use OUT_PWD.
Task-number: QTBUG-57120
Change-Id: I69629bf497931574bff8452939170abb1776ab60
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
60985aa42 converted the QT_CONFIG use, not taking into account that the
feature isn't actually known to the configure system - it's coming
directly from the makespec. so revert that hunk (until we have a better
integration between makespecs and configure).
Task-number: QTBUG-57039
Change-Id: Iaf57b5f5339250055f1c378e091da3ab3fcd4292
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
We can not improve the result from JSON parsing without changing API,
so instead recalculate the line and column based on input and offset.
Change-Id: I54149233f71023aa5d30deff854d6f3406c5c48c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Do not consume white-space after a token before the token has been
parsed, otherwise we end up with misleading offsets. This also fixes
a wrong error of illegal number in several cases.
Change-Id: I492ca4de0346a1d0ab73b1c23d7a72dba812664c
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The old code was wrong as it used QString::fromUtf8() to convert the
char * array to a QString. This could lead to wrong positions when trying
to find the #version directive, as any valid utf8 sequence in a comment
before could have lead to wrong offsets compared to the 8 bit data.
Fix this and optimize the code at the same time to avoid creating a copy
of the data and using QRegularExpression.
Also fixes building Qt with -no-feature-regularexpression
Change-Id: I2d84875ef59b3a0fb4d8b069bf56f4372c57ccdc
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Returning NSNotFound from the NSTextInputClient selectedRange
implementation when there is no selection prevents
dictation from activating (for unknown reasons).
Return an empty {0, 0} range instead. Text input
methods such as Pinyin still work after this change.
[ChangeLog][macOS] Speech to text dictation now works
for Qt text input.
Change-Id: Ibf1729bdd271e8ed5ce3c9d2a0373c8ab3613d8e
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
In XCB environment the QClipboard::changed() was not delivered if the
QClipboard::clear() was issued by other Qt app/process.
If the QClipboard::clear() is used, then the owner in
xcb_xfixes_selection_notify_event_t is XCB_NONE, so we need make the
decission to handle this event by the selection_timestamp and our
m_timestamp[mode].
Task-number: QTBUG-56972
Change-Id: If4c486ac02223eac506465cac7ff1a07bd02a187
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Anton Kudryavtsev
Friedemann Kleint
Marc Mutz
Simon Hausmann
Kai Koehne
Mitch Curtis
Lars Knoll
Robin Burchell
Gabriel de Dietrich
Maurice Kalinowski
Liang Qi
Oswald Buddenhagen
Allan Sandfeld Jensen
Morten Johan Sørvig
Palo Kisa