It doesn't seem to be useful and wasn't really specified in 2 out of 3
existing calls to this function and was probably wrongly specified in
the remaining one, so just remove it for now, it can always be added
later if we decide what exactly should it do.
Only assign known good values of the index to "indexLast" as otherwise
we could end up returning an invalid index value from it if "indexBest"
was never set and the last index wasn't a valid one, as it happened in
wxBitmapBundleImplSet containing an original bitmap and a x2 scaled copy
of it, which shouldn't be used for further scaling.
Add a unit test checking that this bug is fixed now.
Always rounding non-integer scale when upscaling is wrong, as we could
be able to upscale an existing x1.5 image to e.g. 300% DPI scaling with
relatively good results, so only do it if there is no available scale
that exactly divides the requested one.
Add a previously failing unit test which passes now.
It's not necessarily the largest bitmap, but one which may be scaled
using an integer factor.
Fix the tests to expect the correct results, now that they actually
pass.
Check not only that the resulting bitmap has the right size, but also
that it was obtained by scaling the correct bitmap.
Define the required machinery to allow using CATCH CHECK_THAT() macro to
succinctly verify both.
Note that this doesn't change the existing tests, but just rewrites them
in a new form and also adds more tests for a bundle with 3 bitmaps.
Doing anything else results in really poor results, so don't even try
scaling images by 2.25, for example -- scaling by 2 looks much better
and the size difference is relatively insignificant and definitely not
worth the difference in quality.
The changes of 50bc3ceb04 (Get rid of CppUnit boilerplate in atomic unit
test, 2022-05-11) resulted in warnings about possibly uninitialized
variables in this test, so do initialize them for MSVC even if it isn't
really necessary.
Don't use external image files, this makes it simpler to run the test
(it can be now done from any directory) and also experiment with it
(e.g. by making the test image fully transparent or fully opaque).
Remove the now unneeded .ico file but keep the .png one still used by
another test.
Unlike direct tests of __cplusplus, using this macro also works with
(recent enough, i.e. MSVS 2015.3 or later) MSVC versions, even if
/Zc:__cplusplus is not used.
This simplifies some checks and makes some other ones (notably the check
for C++20 used before wxALLOW_COMBINING_ENUMS macro definition) work
with MSVC versions with C++20 support as intended.
This is easy to do and can be useful when using an atomic initialization
counter.
The only platform where not returning the new value might be more
efficient than returning it is Solaris which is not really relevant any
more and on all the other platforms the functions we were already using
provided this value already (or almost, in case of using gcc builtins,
when we just have to use a different one).
Don't take the value of "mask" parameter of wxImageList constructor too
prescriptively, it predates support for alpha in wxWidgets by many years
and was never meant to actually suppress using it.
Instead, do the best thing we can in all cases, i.e. use alpha if it's
specified and supported and use mask otherwise. But only create the mask
from light grey colour if we have nothing else if "mask" is true in
wxImageList constructor, as this is a potentially destructive action
that must not be performed if the user has explicitly decided to set
this parameter to false.
Incidentally fix handling of alpha with comctl32.dll v5 (which is used
in the absence of any manifest) by converting it to a mask in this case:
this is not ideal, but better than just using black background as it
happened before, and restores pre-3.1.5 behaviour.
Also simplify the generic version which just has to create the default
mask if necessary and doesn't have to do anything at all in all the
other cases.
Note that these changes required relaxing some of the existing unit
tests as wxMSW implementation now can add alpha channel to the bitmaps
that didn't have it -- but this is a more useful behaviour, and so it
makes more sense to adapt the tests to it rather than doing a less
useful thing just to conform to the tests.
This commit is best viewed with git --color-moved
--color-moved-ws=ignore-all-spac options.
Closes#22349.
Debugging confirms that DPtoLP() simply returns wrong (i.e. different
from that returned under actual MSW) result when using Wine, so just
account for it in the test as it seems to be better than just skipping
the test entirely under Wine and there doesn't seem to be anything else
to do.
Add wxSharedClientDataContainer class storing ref-counted client data
and use it instead of plain wxClientDataContainer in wxGridCellAttr,
wxGridCellEditor and wxGridCellRenderer classes.
This allows to keep the same client data associated with many grid cells
without having to make many copies of it.
Fix the bug introduced in ec0734f96f (Install DLLs in bindir, not
libdir, when using MSW toolchains, 2021-01-09): the directory where the
DLLs were installed wasn't created any more, resulting in errors if it
didn't exist.
Update to latest bakefile version adding the missing mkdir command to
fix this and also use a released bakefile version for the wx makefiles.
Regenerate configure to match the new version.
See #14601.
The name resolves to Cloudflare proxy which redirects all HTTP URLs to
HTTPS, which breaks the existing tests using HTTP, so prevent this from
happening by using the actual IP address instead.
When using Cloudflare, as we do for www.wxwidgets.org, the connection
succeeds as long as Cloudflare itself works, but reading later fails if
the real server behind Cloudflare proxy does not, so check if we can
read something from it.
Get rid of the ugly WX_ASSERT_MESSAGE() and also use REQUIRE/CATCH
directly instead of CppUnit compatibility macros.
Also use sections to execute the next loop iteration(s) even if the
current one fails.
This commit is best viewed ignoring whitespace-only changes.
The changes of 2144ca38d2 (Get rid of CppUnit boilerplate in
DynamicLibraryTestCase, 2022-04-17) accidentally undid the workaround
from 054cb35b39 (Workaround for a crash with gcc 9.1 from MSYS2 MinGW
32bit, 2019-08-03), so work around the same problem again by avoiding
using CHECK() with function pointers.
Use CATCH macros directly, this is simpler and more clear.
Also use narrow strings instead of wide ones and get rid of wxT() too.
No real changes.
This commit is best viewed ignoring whitespace-only changes.
Under Monterrey (macOS 10.12) we can actually load the file using
wxDynamicLibrary even though it does _not_ exist on the disk, so don't
skip the test just because of this.
Make this function virtual in the base class so that it could be
overridden to do the right thing in wxAuiNotebook, instead of just
always returning NULL as before and add a unit test checking that it
works.
Explain that wxBookCtrlBase::m_pages may not be used in the derived
classes, but that in this case they must override all the methods using
it.
Finally, "soft-deprecate" wxAuiNotebook::GetPageIndex(), which is
identical to FindPage() now.
This fixes infinite recursion when handling wxEVT_HELP in wxAuiNotebook
in wxUniv too, see #22309.
Closes#15932.
After the fixes to UNC paths handling, Normalize(wxPATH_NORM_LONG)
started adding an extra colon after the UNC paths.
Fix this by using wxGetVolumeString() instead of constructing this thing
manually and add a unit test checking for this.
Closes#22275.
Using LogEnabler in wxFileSystemWatcher test case still didn't show
anything even in case of unexpected failures because debug messages
didn't appear in the test output.
Fix this by using a custom log target which shows the debug (and trace)
messages on stderr, instead of using the debug output for them, even
under MSW.
Also make LogEnabler public, and rename it to a more unique name, as it
could be useful in the other tests too.
This ctor was previously present only in wxMSW, make it available in all
ports to allow the same code to compile everywhere.
In most of them wxDC argument is simply ignored, but in wxGTK and wxOSX
it is used to assign the appropriate scale factor for the new bitmap.
Enable previously wxMSW-only unit test checking for this.
This is similar to db6d565fad (Handle wxDC scale factor in wxBitmap ctor
taking wxDC in wxMSW, 2022-04-10), but for the constructor taking
wxImage and wxDC -- it should also use the DC scale factor, even if it
is not used for scaling the bitmap size in this case.
Add a unit test checking for this.
This is similar to 7382e99bbb (Create bitmaps even when given a wxDC
without an HDC in wxMSW, 2022-04-10) and does the same change to
wxBitmap ctor from wxImage and wxDC -- and also extends the unit test to
check for this under wxMSW, which is the only port in which this ctor is
currently available.
