With this MR, the sidebar can now be (if not locked):
* docked to the left or right side
* undocked and floated as an independent window
* closed with the close button in the header
BUG: 455013
When using the windows dark theme it's annoying to have white titlebars on some apps. I noticed that this app which and I would like to use on Windows but the lack of dark theme on the titlebar makes it not fit in well. I did some research on this and found that the best way to set it seems to be using the darkmode toggle on the platform. That can be set as a parameter or environment variable. This seems to be the easiest way to set it. I tested this locally on a Win11 VM and it looks to work the same as setting -platform windows:darkmode=1 or setting QT_QPA_PLATFORM=windows:darkmode=1. Note this does need to be set before the QApplication is originally created and so I think it has to be done per app to work. If I missed some other way to set this that would be better I would be interested to know more.
QT Docs: https://doc.qt.io/qt-5/qguiapplication.html#platform-specific-arguments
It turns out that there were two issues at play here: rounding errors
meant that pixmaps were almost never the same size as pagePainter
thought the tiles should be, and the tile-is-the-size-it-should-be code
path was broken (but only hit in rare cases, seemingly at random).
To help with rounding errors in the future, I added a geometryF function
to NormalizedRect that returns a QRectF. In general,
device-independent-pixel points/rects should be floating point, and
device-pixel rects should be integer.
PagePainter::paintCroppedPageOnPainter had a few calls to
QPixmap::setDevicePixelRatio on pixmaps it does not own, which detaches
the pixmap, making a deep copy. It turns out these were all unnecessary.
It also copied scaled before drawing them onto the painter.
I've tested all the changes except for the annotation stamps, which I'm
just assuming works like the others.
Also ran clang-format.
There are three reasons why scrolling happens (first two cases of scrolling are actually due to
viewport transition that is triggered by annotation being traslated/created):
1. VisiblePageRect are NormalizedRect and should be represented by four values in [0, 1] interval.
Due to rounding errors caused by using QRect, they only reach the [0, 0.999] interval.
That means that right and bottom edge of the page are never visible so if annotation
is translated along the right/bottom edge viewport transition occurs (fixed in
PageView::slotRequestVisiblePixmaps)
2. For freehand line there are some constants (dX, dY) that make a line to be drawn outside
the page borders causing viewport transition. Removing these constants solves the problem (fixed
in SmoothPathEngine::event)
3. Browse mode is activated as soon as the cursor is moved outside the viewport.
Instead, we should forward the move event to the annotation which than can ignore the event (fixed
in PageView::mouseMoveEvent)
BUG: 436742
Currently, scrolling through a document using the touchpad is quite slow compared to scrolling through the standard Qt controls. This MR increases the speed of scrolling with the touchpad by removing the "magic" divider.
CCBUG: 455014
l10n daemon script
Eugene Popov
Nicholas Omann
Albert Astals Cid
Max Mueggler
Paul Worrall
Phu Nguyen
Heiko Becker
Nikola Nikolic