| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
| |
The LoadImage function doesn't seem to work for loading cursors from
resources. Also, it isn't marked as deprecated for Windows CE like on
desktop Windows. So we'll just use it again.
Reviewed-by: thartman
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch includes lots of refactoring, but the real problem was that
in QWidgetPrivate::paintBackground we call drawPrimitive(PE_Widget) with
a potentialy translated painter, but the opt.rect is not translated.
When having a scroll area the calling function used to translated the
painter and then pass the offset around to rectify. but drawPrimitive
cannot rectify it.
The solution is not to translate the painter but use other way to
rectify the brush
Task-number: 257517
Reviewed-by: bnilsen
|
|
|
|
| |
when only setting a border with the stylesheet
|
|
|
|
|
|
|
|
| |
This is Windows CE only. For Windows mobile we don't support context
menus for system tray icons.
Task-number: 250528
Reviewed-by: thartman
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
iteration before the views are updated.
A full scene update (scene.update()) already supported it because the
scene called update() on the views directly. However, partially scene
updates (scene.update(rect)) required two event-loop iterations
before the views were updated.
Auto-test included.
|
| | |
|
|/
|
|
| |
This obsoletes QGraphicsView::handlesChildEvents.
|
|
|
|
|
|
|
|
| |
Qt::TextShowMnemonic removes the '&' for us.
This is required if we want the multi-lenght-variant strings working.
Reviewed-by: Thierry
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On the mac, we were setting the absolute path of the file passed in
while we document that we only set the filename. The filename is
actually what all the other Mac apps do, so we should really do it.
Unfortunately, this task laid around longer than it should, but the
description text was bad.
Additionally, fix a bug in the Cocoa version where we would set a proxy
icon for items that didn't exist, this is not a good idea.
Task-number: 220501
Reviewed-by: Trust Me
|
|
|
|
|
|
|
| |
Found while documenting this internally.
I guess I had a reason for it at one point, but it
seems that I'm more concerned about what index. So
just get rid of this.
|
|
|
|
|
|
| |
These functions hide all this horrible casting and
we've had them for a while. When I see them, I try
to replace them as it keeps the code much more readable.
|
|
|
|
|
| |
we should include qt_windows.h and not windows.h because we have to
define WINVER to 0x500.
|
| |
|
|
|
|
|
|
| |
Merge-request: 653
Reviewed-by: Olivier Goffart <ogoffart@trolltech.com>
Task-number: 247560
|
|
|
|
|
|
| |
Task-number: 248429
Merge-request: 786
Reviewed-by: Olivier Goffart <ogoffart@trolltech.com>
|
|
|
|
|
|
|
| |
for roman numbering of lists as supported by HTML/ODF
Reviewed-by: Olivier Goffart
Merge-request: 681
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This has always been a bit bumpy, the problem is that the popup normally
has its own styling from the desktop, plus it's its own top-level and
that is normally a boundary for propagation. Of course, people are
surprised by this (especially when it works for editable). So, we need
to be a bit better propagating the info. Also the QStyleOptionMenuItem
has the correct font, but if it's set on a window, by the time it
reaches the popup, its resolve mask is very weak, so it will fail to
resolve at all. Setting the point size allows the font to have a bit of
strength.
Task-number: 257486
Reviewed-by: Jens Bache-Wiig
|
|
|
|
|
|
|
|
|
|
| |
QGraphicsScene::sceneRect() returns the growingItemsBoundingRect if not
a particular scene rect is set; otherwise the sceneRect is returned.
This cut-off aims to do less processing when the exposed region covers
the entire scene, but it won't get hit in case of a sceneRect because we
always check against the growingItemsBoundingRect.
Task-number: 257192
|
|
|
|
|
|
|
| |
QSyntaxHighlighter::rehighlightBlock()
Merge-request: 379
Reviewed-by: Simon Hausmann <simon.hausmann@nokia.com>
|
|
|
|
|
| |
Merge-request: 379
Reviewed-by: Simon Hausmann <simon.hausmann@nokia.com>
|
|
|
|
|
|
|
|
|
|
|
| |
NT support"
tst_QGraphicsProxyWidget crashed because the QAlphaWidget tried to access a
deleted widget. Before we had the if check, but that was removed
with this commit: 55137901. Completely wrong, we must check the widget pointer
before using it.
Reviewed-by: jbache
|
|\
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
src/plugins/kbddrivers/usb/main.cpp
tests/auto/qnetworkreply/tst_qnetworkreply.cpp
tests/auto/qwidget/tst_qwidget.cpp
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The inputContext's focusWidget was not reset when disabling input
methods.
Thanks to Benjamin P.
Task-number: 257832
Reviewed-by: Denis
|
| | |
|
| |
| |
| |
| | |
It is useless to store the vector of modelindex from intersectingSet
|
| |
| |
| |
| | |
On windows it makes it 2x faster
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
QListView know exactly what they have on their viewport and we only
paint items clipped to the viewport. So we don't need to ask for each
item its visualRect.
NB: QTreeView and QTableView probably deservee the same treatment
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | | |
Conflicts:
src/gui/graphicsview/qgraphicsitem_p.h
tests/auto/qgraphicsproxywidget/tst_qgraphicsproxywidget.cpp
tests/auto/qgraphicsview/tst_qgraphicsview.cpp
|
| | |
| | |
| | |
| | | |
Task-number: 233342
|
| | |
| | |
| | |
| | |
| | |
| | | |
On 64-bit an id (void *) is 64-bit also, so, it really should be a
pointer, but I'll make it a 64-bit int for the time being just so stuff
compiles.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We don't need to draw all the items that are selected. We just need
those whose rect intersects the one from the viewport.
Task-number: 233342
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This is basically the Windows version of the bug fixed in change
82e825ed841bce324a6892fcbace03f9936d4f4f
Merge-request: 855
Reviewed-by: Norwegian Rock Cat <qt-info@nokia.com>
|
| | |
| | |
| | |
| | | |
Task-number: 248688
|
| | |
| | |
| | |
| | | |
Task-number: 240266
|
| | |
| | |
| | |
| | | |
The layoutState is already current (ie. already applied).
|
| | |
| | |
| | |
| | | |
Task-number: 257579
|
| | |
| | |
| | |
| | |
| | |
| | | |
Maximum number of decimals is DBL_MAX_10_EXP + DBL_DIG
Task-number: 257291
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Most Wacom tablet have a coordinate origin at 0 (Bamboo,Intous), but
some tablet (like DTF 720, which have an integrated screen) have a non
zero coordinate origin. Which lead to an errounous y/a tablet pos
reported by Qt tablet event.
Merge-request: 822
Reviewed-by: Norwegian Rock Cat <qt-info@nokia.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Basically if you would hide a toolbar in the unified toolbar, you
would still see a little bit of area at the top instead of having
everything flush with the titlebar. This change basically unsures that
the unified toolbar makes a decision to hide itself if all the toolbars
inside it are hidden. It makes the behavior of clicking on the toolbar
button behave more or less correctly since we are going to show the
unified toolbar whether we want to or not.
This all will get the toolbar button switch event to be dispatched in
Cocoa as well.
Also add an optimization for checking if we need to change the geometry.
If we don't have any items the other toolbar areas, we can skip the set
geometry call, which wrecks havoc with things in Cocoa.
We still don't solve the case of someone who has hidden the items with
the toolbar button then goes full-screen, then goes back out. I'm not
motivated to solve it as is because we need to keep track of the
hides we do on the button press vs. other hides from the user and still
people can workaround it easy enough by handling window state change and
doing what is recommended in the docs.
Task-number: 208439
Rev-by: Denis
|
| | |
| | |
| | |
| | | |
Reviewed-by: Thierry
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | | |
QTimeLine is now no more used in private APIs
|
| | |
| | |
| | |
| | | |
currentChange is slot in the public class (QAbstractItemView
|
| | | |
|
| | |
| | |
| | |
| | |
| | | |
Triggered on Designer startup on Linux.
Acked-by: Thierry Bastian <thierry.bastian@nokia.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Using qMemCopy limits QGenericMatrix to plain old types like
float and double. Copying the values normally will allow
copy constructors to work on non plain types.
Reviewed-by: trustme
|