Apparently notepad can't be killed by sending WM_QUIT to its process
under Windows 11 because the process we launch is not actually the GUI
process showing the notepad window but some helper parent process.
Work around this by launching mspaint, which can still be killed by
wxKill(), instead.
Closes#23365.
- Update parsing IPvX addresses to follow RFC and add many more tests.
- Rework authority parsing for RFC edge cases.
- Fix a couple of other extreme edge cases.
Closes#23360.
Using sleep was not only fragile but also resulted in tons of TSAN
errors, so replace it by the similar approach to the one which was
already used to wait for the threads to start up.
This is horribly inefficient but we don't care about this in the test
and, at least, this does ensure that the threads exit before it ends.
Don't use sleep in this test as the thread may terminate before it
returns, possibly resulting in crashes that the existing comment warned
about, and definitely resulting in TSAN errors.
Also ensure that the thread doesn't terminate instead of just hoping
that it doesn't do it because it doesn't have time to run 30 iterations
of its entry function by passing the special value of 0 to ctor, which
explicitly prevents the thread from terminating on its own.
Improvements to wxApp initialization and modernize the code to use
std::unique_ptr<> instead of wxScopedPtr or, worse, macro-based scoped
pointers.
See #23319.
While these array classes are deprecated in the user code, they're still
used, for compatibility, in many places in wxWidgets API and allowing to
create them from initializer_list makes using it more ergonomic as it's
now possible to just pass an initializer list of items to fill the
control with, for example, instead of appending them one by one.
Closes#23309.
In addition to the decimal separator, show the date format for the
current locale too in wxUILocale::ShowSystem pseudo-test to check that
this works as expected.
Add more static casts to avoid harmless warnings about long and/or
size_t to unsigned int conversions and enable -Wshorten-64-to-32 in the
allheaders test to ensure they don't get introduced again in the future.
The existing test with this name only checked wxFlexGridSizer::Layout()
and the base class uses a different implementation, which merits its own
unit test.
Setting wxFLEX_GROWMODE_ALL for the non-flexible direction broke
handling of proportions in the flexible direction which were
unexpectedly not taken into account at all any more due to not passing
them to DoAdjustForGrowables().
Fix this by still respecting the proportions in this case and add a test
case to ensure this doesn't get broken again.
Closes#23251.
See #23253.
Don't always "succeed" in wxUILocale::UseDefault() and, consequently, in
wxLocale::Init(wxLANGUAGE_DEFAULT), under Unix systems, even if the
locale couldn't actually be set, as it can happen if the environment
variables contain a locale which is not supported on the current system,
e.g. a "mixed" locale such as "en_FR", or even a completely invalid
string such as "bloordyblop", which still used to succeed.
For now only fix it for reasonably modern systems with locale_t support,
it could be done even for the ancient ones without it later too if
anybody still cares about them.
Closes#23218.
This function replaces the existing GetSystemLocale() as it can
represent the locales that don't have any corresponding wxLanguage
values and for which GetSystemLocale() has no choice but to return
wxLANGUAGE_UNKNOWN.
When looking for the language information, we must recognize the
language independently of the region it is followed by, so en_FR is
still English and fr_DE is still French, even if the full locale is
unknown, but this wasn't the case before.
Fix this by comparing the language part of wxLanguageInfo with just the
language of wxLocaleIdent we're trying to match, instead of comparing it
with its full BCP47 tag, which is never going to match.
At least under macOS it fails when the system locale is something like
en_DE, for example, as setlocale() doesn't support such locales.
Use wxUILocale::GetSystemLanguage() instead of wxLANGUAGE_DEFAULT.
This can be useful to compare the results of calling
wxUILocale::GetSystemLanguage() and the function with the same name in
wxLocale (which actually corresponds to wxUILocale::GetSystemLocale()).
wxGraphicsContext::ResetClip (wxOSX)
Using native CGContextResetClip() to reset the clip causes some
drawing glitches - especially with drawing text with CT - see
https://github.com/wxWidgets/wxWidgets/issues/22629#issuecomment-1399554002
For the time being we have to revert to the previous code
with emulated resetting based on restoring the graphics context state.
We also have to disable a unit tests that fails with emulated
resetting the clip.
See #22629.
This is a generalization of the existing "platform" attribute and allows
to only include some XRC elements if and only if an arbitrary "feature"
is enabled by the application before loading the XRC document.
Work around a bug in wxOSX which breaks the test if we call DumpFont()
because this function calls wxFont::GetPixelSize() which actually
changes the size of the font if it was specified in pixels.
This should be reverted once the bug is fixed.
See #23144.
In wxMSW wxFont::GetPixelSize() returns the size that the font was
originally created from while in the other ports it returns the actual
size being used.
wxMSW behaviour is probably wrong and should be changed, but for now
just restrict the comparison to the vertical component to avoid the test
failures.
Such functions modifying errno is undocumented, and may come
as a surprise to the caller. Consequently also a call to
wxString::Format() would modify errno.
Drop disabled by default and pretty useless memory tracing code and all
the overlapping and poorly documented build options related to it.
Remove memory.cpp entirely and update all the make/project files, but
preserve the now completely trivial wx/memory.h for compatibility and
also keep including wx/string.h from wx/object.h as it seems like a lot
of existing code actually depends on this, even if it should not.
Replace the options in the propgrid sample with a couple of other
debug-related options that were not used before to avoid leaving the
"Debugging Section" completely empty.
With delay after refreshing testing window added in 808bbe8341 ("Fix
executing wxPaintDC-related tests under wxGTK", 2023-01-04) wxPaintDC
tests work also under wxOSX.
