| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Fixes #9191
|
|
This seems like an oversight, that we'd replace ppc with ppc64 on AIX
for the cpu_family, but not for the specific cpu.
|
|
Same thing, but for the more specific cases
|
|
This should help prevent regressions.
|
|
The correct way to mark these is `KwargInfo(..., (T, type(None)))`.
There's also a few cases of `(T, None)` which is invalid, as `None`
isn't a type
|
|
This is quite valuable for enum-like inputs, where only a certain set
of values is allowed.
|
|
It should build the fallback subprject with default_library=static and
override the dependency for both static=True and static kwarg not given.
Fixes: #8050.
|
|
using `TestCase.addCleanup` is both neat and tidy, and also ensures that
the cleanup will be done, even if the test fails.
|
|
Fixes: #7007.
|
|
This adds a full set of `typed_pos_args` and `typed_kwarg` decorations,
as well as fixing all of the typing errors reported by mypy.
|
|
This adds a new keyword argument to the init method, `allow_fail`. When
set to True (default is False) then a failure to configure is not an
error, and output is still returned. This can be useful for cases where
we expect initialization to fail, and want to check the output.
|
|
There are two problems with having this in the try/except block. The
first is that both of the if statements will raise, and the except
statement cathces `Exception`, so it catches these two cases, prints a
message that we either don't want or already printed, then re-raises.
|
|
|
|
and clean up all outstanding issues
Skip 'test cases/common/141 special characters/meson.build' since it
intentionally uses trailing newlines.
|
|
We don't actually want to do anything with the open()ed file, just
immediately close it.
The CalledProcessError doesn't have its return returncode checked
here, even though other code with the same type of context manager does.
|
|
This reverts commit 0b97d585480e973d8b149618901f7a4ddfa1a906.
|
|
Otherwise it always returns the value for c++98, starting with MSVC 2017
15.7 or later. Earlier versions are not affected by this mis-feature
|
|
In the case main->subp->subsubp, if subsubp succeed to configure but
subp subsequentially fails, subsubp is still being built but its summary
was missing.
|
|
fixs: #9000 Meson not correctly process with -l:xxx.a link arguments in pkgconfig .pc file.
see also:https://stackoverflow.com/questions/48532868/gcc-library-option-with-a-colon-llibevent-a
with unit test, unit test will be partially skiped if pkg-config version < 0.28 .
see: https://gitlab.freedesktop.org/pkg-config/pkg-config/-/blob/master/NEWS
|
|
Fixes: #9065
|
|
Gcovr interprets exclude filters, as passed to the -e option, as
regexes. Since we want to exclude a raw path, the argument must be
escaped.
|
|
Meson already works like that, except in do_copydir() that requires
absolute destdir. Better explicitly support that instead of leaving it
undefined and unconsistent.
|
|
AS we've run into a few programs using ansi escapes in their version
outputs, we'd like to test them as well.
|
|
|
