The code would create a dictionary that was of type `str : list[str] |
str | None`. Then would later try to call `len(' '.join(dict[key]))`.
This would result in two different bugs:
1. If the value is `None` it would except, since None isn't iterable
and cannot be converted to a string
2. If the value was a string, then it would double the length of the
actual string and return that, by adding a space between each
character
This allows tests to check whether stdin is a tty to figure out if they're
running in interactive mode or not.
It also makes sure that tests that are not running in interactive mode
don't inadvertendly try to read from stdin.
This is very similar to --gdb, except it doesn't spawn GDB, but
connects stdin/stdout/stderr directly to the test itself. This allows
interacting with integration tests that spawn a shell in a container
or virtual machine when the test fails.
In systemd we're migrating our integration tests to run using the
meson test runner. We want to allow interactive debugging of failed
tests directly in the virtual machine or container that is spawned
to run the test. To make this possible, we need meson test to connect
stdin/stdout/stderr of the test directly to the user's terminal, just
like is done with the --gdb option.
When getting debug file arguments we can sometimes pass None, where a
None is unexpected. This becomes a particular problem in the Cuda
compiler, where the output will unconditionally be concatenated with a
static string, resulting in an uncaught exception. This is really easy
to spot once we annotate the functions in question, where a static type
checker like mypy easily spots the issue.
This commit adds those annotations, and then fixes the resulting error.
Fixes: #12997
Any code that needs to know mesonlib.python_command currently assumes
the PyInstaller bundle reference is added to the sys module, which means
that it assumes the only freeze tool used is PyInstaller. Really, all we
need to check is sys.frozen as it's never normally set, but always set
for a freeze. We don't care if it was PyInstaller specifically, and we
don't need its bundle directory.
See https://github.com/mesonbuild/meson/discussions/13007
Popen_safe_logged has a small inefficiency. It evaluates the stripped
version of stdout/stderr before checking if it exists, for logging
purposes. This would sometimes crash, if it was None instead of ''.
Fixes#12979
It is generally accepted practice to convert dict.keys() to a list
before iterating over it and e.g. deleting values, as keys returns a
live-action view. In this case, we use the difference of *two* dict
keys, which returns a regular non-view set and doesn't need protecting.
Iteration order doesn't matter (the set already randomizes it anyway).
Avoid the cost of converting to a list.
Only Environment and ConfigurationData are mutable. However, only
ConfigurationData becomes immutable after first use which is
inconsistent.
This deprecates modification after first use of Environment object and
clarify documentation.
We have seen a number of bugs from people confused by warning that the
need both a C and C++ compiler to use the CMake method. This attempts
to provide a more helpful error message.
When a user writes `import'foo')`, Meson checks the
`mesonbuild/modules/` directory for a package called `foo.py`, and
attempts to import it. We don't want to expose any implementation detail
packages like `_qt.py`, so if someone write `import('_qt')`, we should
immediately give a "doesn't exist" error.
Since it's an implementation detail, and shouldn't be exposed. This also
helps give better error messages when a user writes `import('qt')`,
since otherwise you get an error about `qt doesn't have an initialize
method`, and now you get `qt module doesn't exist`.
Commit 83facb3959 switched to using
`textwrap.dedent` for the code templates for `gnome.mkenums_simple`.
That changed indentation, however, making the generated code harder to
understand.
We improve this by properly indenting the multiline strings before
dedenting them. For optional parameters `decl_decorator` and
`header_prefix`, we add a newline if they are set to keep separation
between generated code blocks.
The `AstVisitor` intentionally ignores whitespaces and symbols,
as they are not useful for tne interpreter. However, when formatting a
build file, we need them. This commit introduces a `FullAstVisitor` that
visits every Nodes, including whitespaces and symbols.
This incorrectly warns that `compiler.run()` is new in 1.5.0 for
Fortran, but that works fine for older versions (noted with 1.2.3). This
makes sense, as FortranCompiler inherits CLikeCompiler
If -L flags get into CLikeCompiler::build_wrapper_args, they will be
correctly detected and the /LINK flag added to the list. However,
CompilerArgs::__iadd__ will reorder them to the front, thinking they're
GNU-style flags, and this will cause MSVC to ignore them after
conversion.
The fix is twofold:
1. Convert all the linker args into their compiler form, making sure the
/LINK argument is dropped (see 2)
2. Insert /LINK into extra_args if not already present
3. Execute in situ the unix_to_native replacement, ensuring no further
reordering occurs.
Fixes#11113
This somewhat aligns "darwin" (Mach-O) with how ELF RPATHs are treated.
Instead of blindly removing all RPATHs, only remove the ones that are in
the rpath_dirs_to_remove set. This way, RPATHs that were added by the
toolchain or user are left untouched.
It is important not to remove RPATHs as they may be vital for the
executable at runtime. Issues #12045 and #12288 are examples of this.
Issue: https://github.com/mesonbuild/meson/issues/12045
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
Signed-off-by: Eli Schwartz <eschwartz93@gmail.com>
There's a known ninja bug
(https://github.com/ninja-build/ninja/issues/1952) that running this
with dyndeps will result in Ninja deleting implicit outputs from the
dyndeps, leading to pointless rebuilds. For reference, this is what
CMake does as well.
This basically existed for an assert which we don't need, as mypy would
catch that issue for us anyway. Removing the function entirely has some
small performance advantages
We already have to decide whether to scan a file at configure time, so
we don't want to have to do it again at compile time, every time the
depscan rule is run. We can do this by saving and passing the language
to use in the pickle, so depscan doesn't have to re-calculate it. As an
added bonus, this removes an import from depscan
We don't need to write and pass two separate files to the depscanner,
I've used the pickle because the pickle serializer/deserializer should
be faster than JSON, thought I haven't tested.
Dylan Baker
Jouke Witteveen
Daan De Meyer
Ben Corby
Tristan Partin
Momtchil Momtchev
Ole André Vadla Ravnås
Jonathon Anderson
Eli Schwartz
William D. Jones
Xavier Claessens
L. E. Segovia
Matthijs Velsink
Charles Brunet
Otto Moerbeek
Gerion Entrup
Junior Rantila
Fabian Groffen