Widgets demo:
widgets_rtl.png (view on web)
Grid demo:
griddemo_rtl.png (view on web)
AUI demo:
auidemo_rtl.png (view on web)
https://github.com/wxWidgets/wxWidgets/pull/25822
(12 files)
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.
Amazing!
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.
@AliKet pushed 10 commits.
—
View it on GitHub or unsubscribe.
You are receiving this because you are subscribed to this thread.
@vadz commented on this pull request.
Thanks, this looks good and could be merged AFAICS, I'm not sure why is this in the "draft" state yet?
Also, unfortunately I have no idea why did SliderTestCase
start consistently failing with Qt 6.8 just recently, but this is unrelated to this change.
> @@ -117,6 +122,14 @@ class WXDLLIMPEXP_CORE wxQtDCImpl : public wxDCImpl virtual void* GetHandle() const override { return (void*) m_qtPainter; } + // LRT/RTL related functions
Typo:
⬇️ Suggested change- // LRT/RTL related functions + // LTR/RTL related functions
> @@ -3478,11 +3478,13 @@ wxRect wxAuiManager::CalculateHintRect(wxWindow* pane_window, m_frame->ClientToScreen(&rect.x, &rect.y); +#ifndef __WXQT__
I was always suspicious of this piece of code, if it's not needed for wxQt, it seems to confirm that it shouldn't be necessary for the other ports as well... Do you have any idea what could be wrong in/with them to require it?
In src/qt/dc.cpp:
> @@ -462,10 +496,10 @@ void wxQtDCImpl::Clear() int width, height; DoGetSize(&width, &height); - m_qtPainter->eraseRect( DeviceToLogicalX(0), - DeviceToLogicalY(0), - DeviceToLogicalXRel(width), - DeviceToLogicalYRel(height) ); + const wxPoint pos = DeviceToLogical(0, 0);
Just to confirm: this doesn't really change anything, does it? I.e. DeviceToLogicalX()
etc work too?
> + // For correct result in RTL layout, the drawing is done in a non-mirrored + // DC and the mirroring is done in paintEvent() above.
Sorry, I feel dumb, but I don't really understand why do we need to do this. I.e. what is wrong with using RTL layout for the DC itself instead of mirroring it later?
> + default: + wxFALLTHROUGH; +
I'd remove this to get a warning in the (unlikely) event of more elements being added to this enum.
In tests/graphics/clippingbox.cpp:
> @@ -1033,9 +1033,9 @@ static void OneDevRegionRTL(wxDC& dc, const wxBitmap& bmp, bool useTransformMatr return; } -#ifdef __WXGTK__ +#if defined(__WXGTK__) || defined(__WXMSW__)
Could we perhaps add a check for __WXMSW__
below instead of disabling this test entirely for it?
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.
@AliKet pushed 19 commits.
—
View it on GitHub or unsubscribe.
You are receiving this because you are subscribed to this thread.
@AliKet pushed 12 commits.
—
View it on GitHub or unsubscribe.
You are receiving this because you are subscribed to this thread.