ID2D1RenderTarget::PushAxisAlignedClip/PopAxisAlignedClip used to clip the region (with wxGraphicsRenderer::Clip) and ID2D1RenderTarget::PushLayer/PopLayer used to rendering to the transparent layer (with wxGraphicsRenderer::BeginLayer) are non independent but have to be used in the controlled sequences: "A PushAxisAlignedClip and PopAxisAlignedClip pair can occur around or within a PushLayer and PopLayer, but cannot overlap" (and of course finally each Push* call must have a matching Pop* call).
To control the sequence of access to the AxisAlignedClips and Layers there is implemented a wxStack data member holding Clips/Layers parameters which reflects a physical stack of respective Clips/Layers in ID2D1RenderTarget. This way we know in which order to pop and what to pop from ID2D1RenderTarget stack if there is a need to do so.
Closes#17590
When there is no clipping region currently set then current effective clipping region is identical with entire DC surface and therefore DC size should be returned as a region size.
Closes#17013
Graphics renderers accept negative clipping box size but for internal calculations done in wxDCImpl::DoSetClippingRegion() we need to have a box defined by (x,y) at top-left corner and having non-negative size so we need to recalculate parameters if necessary.
While setting a clipping box there is necessary to intersect it either with current clipping region location if such region exists or with wxDC surface extents if no clipping region is set. This way effective clipping box will be always inside the wxDC surface.
Note: Effective clipping box can be an empty region.
See #17013
Because CreateRectRgn() Win API requires that (x1,y1) parameters represent the top-left corner of the clipping box so if a box with negative values of the width or height is passed to wxMSWDCImpl::DoSetClippingRegion() (what means that (x,y) doesn't represent top-left corner) we need to recalculate passed parameters to get the box with (x,y) at top-left corner.
This reverts commit 71dfb3b414 which fixed
appearance of the captions in high contrast mode but broke them for wxGTK.
Ideal would be to make this code work in wxGTK too or, in the worst, case only
use this code for wxMSW, but for now at least avoid breaking wxGTK appearance
by default.
See #16186.
1. Added checking if actual drawing output is really clipped at the edge of the clipping box.
2. Added tests for wxGCDC with GDI+, Direct2D, Cairo renderers under wxMSW.
3. Added new test to check calculations for bounding box set entirely outside DC area.
This improves their appearance especially when they're used as children of
wxNotebook, which uses a different background colour than plain wxPanel by
default.
This generalizes what was already done for wxChoicebook in 6cab632f
(see #12503) to all book controls.
Closes#16878.
Simplify the officialbuild.bat script by only checking for "vcXX" instead of
accepting both them and "vsYYYY".
Also add instructions for using it in the corresponding HOWTO.
Closes#17514.
This reverts 6c40531fb7 ("Make main thread wake
up code more efficient and less error-prone in wxMSW") as, while being more
efficient, the new code doesn't work at all when we're not running the message
loop ourselves as it happens when the user opens a menu or starts resizing a
window because in both cases Windows runs a local message loop dispatching the
messages itself and this message loop doesn't react to our event object being
signalled.
So this approach can't work and needs to be reverted, even if it reintroduces
the danger of overflowing the message queue (see #9053).
Closes#17579.
Disable thread-safe initialization for static local variables in Visual Studio
2015 when XP toolset is used as this results in crashes when using DLLs under
XP, see #13116.
Closes#17403.
Allocating 100px (+ margins) for empty listboxes in wxMSW resulted in a
surprising behaviour when the best size of a listbox became (significantly)
smaller after adding some items to it.
One possible solution could be to ensure that all listboxes are at least as
wide as empty ones, but it seems wrong to insist on always allocating 100+ px
when a listbox could be much narrower.
So try to mitigate the problem by making the empty listboxes narrower by
default which should reduce the chance of them becoming narrower still after
appending some items to them, even if it doesn't completely eliminate it.
Also, as a side effect, this commit replaces non-font-and-DPI-dependent
hard-coded 100px value with a more reasonable value based on text metrics.
Due to an oversight, it wasn't declared as const, making it impossible to
subtract from a const wxTimeSpan object.
Fix this and add a unit test verifying that this compiles and works as
expected.
Closes#17583.
If there is no wxBitmap selected into wxMemoryDC which is passed to wxGraphicsContext then there is raised an assertion warning (harmless in this context but caught and reported by CppUnit). To suppress this message we need to select any bitmap into wxMemoryDC prior to creating a wxGraphicsContext.
Try to make it possible to understand where exactly do the GTK+ errors
appearing in the buildbot slaves output originate from by installing a custom
Glib log handler and prefixing the normal log messages with the name of the
test running when they're generated.
This required a small refactoring of DetailListener which is now always
installed, in order to always have access to the current test name, but still
needs to be explicitly enabled to produce output.
Destroying a window with mouse capture results in an assert, which is
translated into an exception when running the test suite. As this exception is
thrown from wxWindowBase dtor, it results in an immediate program termination
when using C++11 and can also have the same effect even when using C++98 if
this exception is thrown while already handling another exception due to a
test failure.
Try to avoid this by using a "safe" DeleteTestWindow() function instead of
deleting the window directly. Currently this function ensures that the window
doesn't have mouse capture before deleting it, but it could also be used to
check for other things later. Also, this commit only uses this function for
the two controls which do happen to be destroyed with mouse capture currently
(at least when using wxGTK), but it should probably be generalized to all
controls in the future.
This assert was redundant as a similar check is done in the base class dtor.
And while usually this assert is just annoying, instead of being helpful, when
running the unit test suite which installs a custom assert handler throwing an
exception, it is actively harmful as generating 2 asserts during the
destruction of a window with mouse capture results in immediate termination
(even in C++98 mode, unfortunately the whole idea of throwing from dtor is
probably unsalvageable anyhow when using C++11).
Optimizations were turned off for g++ 4.8 to work around compiler bug
resulting in generating infinite loops for the code in this function in
052e598d09, but apparently other versions of g++
suffer from it as well, so just always turn them off until we can be sure
about the version of the compiler in which this bug was really fixed.
There is no way to show the hint without native support in a control with
wxTE_PASSWORD style, so simply ignore them completely in this case.
Closes#17078.
Only use TBSTYLE_AUTOSIZE, adjusting each button to the size it really needs,
for the toolbars with wxTB_HORZ_LAYOUT style as they don't have any uniform
button size anyhow.
Strip mnemonics correctly from the translations including them in parentheses
after the translated menu text, as is apparently common when using CJK.
Closes#16736.
Don't switch locale to Japanese, which might not be supported, but just use
wxTranslations directly because we only need translations, not full locale
support.
Also rename the test method to a more descriptive name.
Chinese, Japanese etc translations use a special style for the menu mnemonics
and append them to the translated menu label in brackets, e.g. the menu label
could have the form of "<translation-of-file> (&F)".
Check for this style of mnemonics in wxStripMenuCodes() too and strip them as
well.
Fix regression introduced by bd388e9827: bitmap
buttons could now be made significantly taller than the text control height if
their bitmap was big enough. Only make buttons taller if they wouldn't be tall
enough on their own, instead of always increasing their height, even if it's
already big enough.
Closes#17576.
This resulted in a crash in GetDefaultEditorForCell() later when GetEditor()
was called from HideCellEditControl() which is itself always called from
wxGrid dtor because GetDefaultEditorForCell() dereferenced m_table without
checking if it was not null any more.
Add the missing check to this function and GetDefaultRendererForCell() too,
for consistency.
In addition, dismiss the cell editor immediately when changing table instead
of doing it at some later time, as it just doesn't make sense to continue
showing it any more as the data it was started to edit doesn't belong to us
any longer.
The loop over the existing selection was buggy and took into account only one
corner of the block instead of the entire block and also skipped some blocks
entirely.
Closes#17572.
Because label is implemented in the control tool as a separate wxStaticText object which exists only if label is non-empty so we need to handle appropriately also the cases when non-empty label is set to empty and vice versa.
Setting a new label for the button tool with TB_SETBUTTONINFO would require to do some additional actions for other items in the toolbar (manual re-positioning items in the control tools, updating stretchable spacers) so it is easier just to re-create the toolbar with Realize().
See #17567
