... instead of using erase in a loop, with quadratic complexity.
Change-Id: I91053ddb58639615e8864e5be20861e7cceb815e
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
... of QLists. The returned list was only ever exactly of
size one or two, and two only if a local QList in the
getGraphParts() function was not empty.
Instead of using a QList to return those QLists, use a
small struct of two QLists. This way, we avoid copying,
memory allocations, and the code becomes a lot clearer,
because we can give the members of the struct meaningful
names, instead of at(0) and at(1).
As a consequence, at the call site, the loop over the
result can now be collapsed into an if.
Saves ~1200b in text size on optimized GCC 5.3 Linux
AMD64 builds.
Change-Id: I5c2a3b3be0edcc418b310551abca51c410aca7c8
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Instead of move-assigning a new QIcon to the default-
constructed members, initialize them directly in the
ctor-init-list. Do this for the other data member, too,
for consistency.
Also use QStringLiteral instead of QLatin1String as
the argument to the QIcon ctor. This is correct, since
the string will never be modified, and saves two memory
allocations.
Change-Id: I9398bbe9d0fd63db9b6cf856dd240ab334364518
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
This is mostly straight-forward, but some things are worth noting:
1. Yes, this is necessary. The noexcept operator looks for noexcept tagging,
not at the contents of the function to determine whether to return true.
The more conditionally-noexcept functions are used, the more important it
becomes that low-level classes are correctly marked noexcept. In that, it
is like constexpr.
2. In accordance with the rules governing noexcept specifications for the
standard library itself, the get*() functions, as well as any function
taking (but not returning) channel values as ints or qreals, or taking
QStrings are not marked as noexcept, since they have preconditions and
thus a narrow contract. Narrow-contract functions should not be noexcept.
All other functions have wide contracts (ie. no preconditions).
3. Any function returning QString can throw (bad_alloc).
Consequently, they, too, are not marked nothrow.
Change-Id: I023356ba1b9a4b057e613cd45380a89b222dc09d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... Just cache it.
Change-Id: Iae70e09233b237d904ab100f46f8dc7ab9ac8e04
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Thanks to gunnar.roth@gmx.net for updating the WinCE patch.
Change-Id: I131e0ee9141efdbd377e096aae0959311a601830
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
There is nothing guaranteeing there is a context current when the
backingstore dtor is invoked for widget windows that do not contain
render-to-texture widgets. In some cases the eglfs compositor's
context is still current, while in other cases (esp. when having
popups and other top-levels) there is none.
To prevent not releasing the backingstore texture and the incorrect
warning about incorrect context, make the correct context current via
an offscreen surface, when necessary.
Change-Id: Id8257650c1ec8cf96910a4f285b779419c3558a8
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
QWidgetBackingStore::sync() has two variants. The widget texture list
logic was only present in one of them. This led to problems on eglfs
in cases when the other variant got invoked. (for instance using the
scroll area in the qopenglwidget example)
eglfs relies on the texture lists's lock status to properly serialize
its somewhat asynchronous built-in compositing mechanism and therefore
is the only platform affected. The patch moves the code to be invoked
from both sync() variants.
Task-number: QTBUG-50668
Change-Id: I4c62987b7bb3cc40f98a4e94447368d2f740dbfd
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
(or trivially marked const) ... by replacing them
with C++11 range-for loops.
Change-Id: I1522e220a57ecb1c5ee0d4281233b3c3931a2ff8
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
... by replacing them with C++11 range-for loops.
To avoid detaches of these mutable Qt containers,
wrap the container in qAsConst().
Change-Id: If086bea06fe26232a7bb99fad8b09fce4dc74c27
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
... by replacing them with C++11 range-for loops.
This is the simplest of the patch series: Q_FOREACH took a
copy, so we do, too. Except we don't, since we're just
catching the return value that comes out of the function
(RVO). We can't feed the rvalues into range-for, because
they are non-const and would thus detach.
Change-Id: I5834620bf82f3442da7b2838363d351a0fb960a0
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
... by replacing them with C++11 range-for loops.
The functions QMakeProject::values(), QMakeMetaInfo::values()
and QHashIterator::value() all return by const-reference,
so they can be passed to range-for without further changes.
Change-Id: Ic3b39ed8ff8cd7a6f287f1aa9d61a1acd67d7aaa
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
... iterate over the container itself instead. Avoids
temporary QList creation as well as the lookup cost
when actually calling value(key).
Change-Id: Icac867c30e63863cfa44a382eedd4d6df2070a59
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
All other colors are hard-coded in here, only the background followed the
palette.
Change-Id: If8033a72f85f953358ffeec4edd80a36b8494c7f
Reviewed-by: Topi Reiniö <topi.reinio@theqtcompany.com>
As of change a094af0017, X11 now joins
other platforms in having a fake screen if no physical screens are
attached, rather than ever allowing QGuiApplication::screens() to
return an empty list or QGuiApplication::primaryScreen() to return
null. This is more consistent across platforms and helps to prevent
bugs in applications which want to use QScreen in some context but
fail to check whether it exists.
[ChangeLog][X11] In case there are no physical screens attached,
QGuiApplication::screens() and QGuiApplication::primaryScreen() will
return a placeholder QScreen object.
Task-number: QTBUG-42985
Change-Id: I0f58afbfe2a034fb9e872a5d1dbdabad2c7dbf9e
Reviewed-by: Błażej Szczygieł <spaz16@wp.pl>
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
- Introduce Qt 5 signal-slot connection syntax.
- Merge MainWindow::createMenus()/createActions()
into MainWindow::createMenus(), removing the need
to store the actions as member variables.
Use QMenu::addAction() for brevity.
- For actions in QActionGroups, carry the Valuator enum
in QAction::data so that the slot to handle the selection
does not need to compare the QAction pointer itself.
- Use a non-modal QColorDialog, so that the user
can change colors more easily while drawing.
- Choose saner shortcut keys: control-Q should not
override the default usage for quitting the application,
and using shortcuts for About dialogs is anyway dubious.
- Improve the example documentation.
Change-Id: I57aaf5f5b885c13a953482dbcc41275dd3d6bff4
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Topi Reiniö <topi.reinio@theqtcompany.com>
Replace the large switch in QStandardPaths::writableLocation()
by a function mapping QStandardPaths::StandardLocation
to the int clsid required by SHGetSpecialFolderPath(). Warn if
SHGetSpecialFolderPath() fails for config location and append
prefixes (cache/application name/organization) only on success.
Change the logic in QStandardPaths::standardLocations() to append
the writable location first, avoiding the prepend().
Task-number: QTBUG-50570
Change-Id: I9d80e83d1ca7af3ea8d3ac2c720ee981b1b2c32a
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
This partially reverts the source and binary incompatible parts of change
d921a9bd15 that made public members in an
exported class private and changed signature in one case.
Task-number: QTBUG-50703
Change-Id: I2719f276256206347d3c27d80a16db34a4ea2888
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
... but they were only marked as bidirectional.
Fixed.
This change is slightly BiC, because if some class used tag
dispatching on this iterator type, a recompile might now
pick a different overload, and the old one may not be
available to a user anymore (no longer instantiated).
I do not think Qt uses that technique, yet, though. Not on
iterator categories, at least.
Change-Id: I75fb334af7e191f882d11575dec83c879a6b50ee
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We depend on the assumption QFontCache::findEngine(key) for key.multi=1
returns a font engine of type QFontEngine::Multi;
guarantee that by checking it in a single place.
Change-Id: I287da4fd62deb22fc5520cde5b0505bc44547609
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
This fixes a regression when entering data in a password field.
The important part is to simply not call convertLineOffset for
single line text edits. The reason is that the function when dealing
with password fields gets an empty string back when calling textAt etc.
This is good since we don't want to leak passwords through a11y apis.
The problem with the functions returning empty strings is that we end up
in an infinite loop in convertLineOffset.
Task-number: QTBUG-49437
Change-Id: I76faa7e33e3ad5c3aeb5c75d8c4b93f1b8227bfc
Reviewed-by: Morten Johan Sørvig <morten.sorvig@theqtcompany.com>
Usually when getting an object from an interface, the object
can be assumed to be valid. We need to check isValid though
since the screen reader access is inherently asynchronous and
objects might be in the QWidget destructor where the QObject is
still valid.
Thus check QAccessibleInterface::isValid in all uses of it in the
OS X implementation.
Task-number: QTBUG-50545
Change-Id: I6e142f6ead1b3281cab2cbc61ce1406bbfe29f69
Reviewed-by: Morten Johan Sørvig <morten.sorvig@theqtcompany.com>
For QLatin1String, operator== is overloaded, so comparing to a latin-1
(C) string literal is efficient, since strlen() is comparatively fast.
OTOH, QStringLiteral, when not using RVO, litters the code with
QString dtor calls, which are not inline. Worse, absent lambdas,
it even allocates memory.
So, just compare using QLatin1String instead.
Change-Id: I761b2b26ab5b416bc695f524a9ee607dacf0a7b2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
When the window container already has a window handle, allow embedded
windows to use this handle as a parent. This change will allow proper
window stacking and clipping.
Task-number: QTBUG-50477
Change-Id: I8d656ecb99e0c42ae7a7ac461e5e5b5d801f5493
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
c7e5e1d9e0 assumed shortcut handling could be done at the QPA
level, but this fails on async platforms where there are e.g
activation events in the QPA queue and we then try to handle
the shortcut synchronously.
This commit restores the handling to QtGui for non-OS X platforms,
and we should fix OS X by adding callbacks through the IME for
the special case of OS X.
Task-number: QTBUG-50467
Change-Id: I7460184cc60d0319f07771eec487325a17d7e7e7
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
- put it in a separate private header
- add a protected QFinalState constructor that takes a
QFinalStatePrivate
- when passing through, also export QStatePrivate, so the linker can
find the symbol
Change-Id: I8563da16739731d9d6142be438a8769fc29de148
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
When we want to use qtTrId function instead QApplication::translate in
generated files it's not possible to use custom function with presented by
the --tr argument due to signature difference between qtTrId and
QApplication::translate function.
With this argument present, uic will generate qtTrId function with second
argument empty. When used together with --tr parameter it will use
presented function instead.
Change-Id: I0d986e4ea25ce58917fd9a2866b1eac0a4671502
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Jarek Kobus <jaroslaw.kobus@theqtcompany.com>
Port a QString::fromLatin1() to QStringLiteral.
The string is never modified.
Wrap char in QLatin1Char to construct a QChar with 8-bit character.
Change-Id: I579ca9eae0c2a72ea893254f258b001063e9f64e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
... instead of using erase in a loop, with quadratic complexity.
Change-Id: I9527923dcfc3201776a59de53dd5c03ca4d80063
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
This test ensure that a connection made to either an invalid host or an
unreachable host fails properly.
Task-number: QTBUG-47452
Change-Id: If31d23d815e496fc21ef2e1e19c34c2723bd3504
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Andy Shaw <andy.shaw@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
By default, the read chunk size is QIODEVICE_BUFFERSIZE and the write
chunk size is 0 (which means that we don't use the internal write buffer).
Derived classes may override these values to define the size of
QIODevice's write buffer or to optimize the read buffer use.
Bump the TypeInformationVersion field in qtHookData, to notify the
Qt Creator developers that the offset of QFilePrivate::fileName was
changed and dumpers should be adapted.
Change-Id: Ib732bc94be8da8a5514a6e5dcc04445895f130d8
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Some sequential devices allow data to be partitioned into several
channels that have the property of independently sequenced delivery.
Supporting such devices uniformly requires a unified API which provides
the user with a uniform concept of multistreaming.
This patch is based on QProcess's multiplexing model and introduces
the following features:
- ability to get the number of channels;
- multiple internal read/write buffers;
- channel selection functions;
- notification signals on channel activity.
To keep the source code compatible with single-channel implementations,
introduce a private class that references the current read buffer and
hides multistreaming internals from the user.
Bump the TypeInformationVersion field in qtHookData, to notify the
Qt Creator developers that the offset of QFilePrivate::fileName was
changed and dumpers should be adapted.
[ChangeLog][QtCore] Added multistreaming to QIODevice.
Change-Id: Idcaa6a618927c101c4c7284d2a633913be6a6ee2
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
The QFontCache instance is stored in QThreadStorage and thus calling
QFontCache::instance() isn't really cheap; avoid calling it multiple
times where possible.
Change-Id: I1b7a83089698a06c49dac08b2a3a8e9c3c75a500
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
Since the likelyhood that decoration_size is greater than 0 then we can
save time with explicitly ensuring that it does not call the functions.
Change-Id: Iaf991a6c77bbc4908c459307ca7e645cf3b3ca20
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
This allows us to support file descriptors >= FD_SETSIZE.
Change-Id: I7e4a35333446a587cfd13c077fa5e19fa3d1abc4
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Instead of building a QMap with dummy values, just to sort the
elements of the QSet, build a QList, and sort that.
Also use QStringList::join() instead of rolling our own loop.
Change-Id: Iebb7faac8e4b72d6f71b3ab3feba7865b1a102f3
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
This fixes the rendering of translucent borders on CSS-styled widgets to
prevent visible overlapping at the corners. This is done by using a
miter joint (45 degree angle) if either of the adjacent edges are
translucent. Previously, adjacent edges would be drawn at full length
and overlap at the corners if both edges are BorderStyle_Solid and have
identical QBrush objects. This works if both QBrush objects are opaque
but causes visible overlap if one or both of them are translucent.
Change-Id: I99d46c8634cb314e642c635439ed2f7819fcba6a
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
The error message was not informative enough to troubleshoot.
Change-Id: Ib21fce238c80b31d21939d139fc173b48eb5ceb8
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Qt 5.7 no longer contains lib/fonts. This is a regression for users
relying on Qt's own fonts when running on Embedded Linux systems without
fontconfig. To alleviate the pain, show a friendly message about this.
The old "is Qt installed correctly" note is out-of-date anyhow now.
Change-Id: Idcb732fb65ed9c45529a0bcc3168227e0969bea0
Reviewed-by: Louai Al-Khanji <louai.al-khanji@theqtcompany.com>
Instead, allocate a temporary buffer on the stack. This prevents the
internal read buffer from being allocated if the device is opened only
for writing.
Change-Id: Ib91c58299206e92006589807527e7b71a5555c8f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
For non-scalable fonts, the units_per_EM in FreeType is
documented to be undefined and will default to 0, which means
that any division by it will cause an exception.
The emSquareSize() function already checks if the font is scalable
and returns y_ppem if not, so lets use it instead in all locations
where we're not already sure the font is scalable.
[ChangeLog][Text][Freetype] Fixed a divide-by-zero exception
when accessing bitmap fonts.
Change-Id: I8839d4c83047fb3f6bb4d69af0258e94a258a4d9
Task-number: QTBUG-45963
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Makes it clear that the engine for font with no Latin WS support
won't be cached as "shared" engine and thus shouldn't be normally
expected in the cache, so don't even try to find it there.
Change-Id: I9e6275b4919607f4057a193b446825c98932bd23
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
