A common, and challenging, issue in CI runners is debugging issues when
you know the information you want to check, but it's in the log file
which you don't have because remote CI machines.
There are various edge cases where this is especially hard to solve,
such as inside of `pip install` where the build directory with the log
file is automatically cleaned up. But it's never really *easy* when you
don't expect it, and the best case scenario is your iteration time gets
cut in half as you hurriedly go add some `cat`s to your CI scripts.
Meson can, at least sometimes, detect platforms where text can be
emitted inside of "folds", which are auto-collapsed and don't obscure
the general output, but when clicked will expand the logfile contents.
Hook this up.
We start off with a Github Actions implementation. We had some internal
code used by our own project tests runner, which can be utilized.
Also permit forcing it via an environment variable, in case
autodetection fails and you just want to force *something*, especially
when meson is called a couple layers deep inside some other tool.
The poorly named `print_tool_versions()` doesn't just print the tools
versions, it finds them and populates a global table, without which some
tests will fail. Rename the function and add a `report` argument so that
calls can decide whether they want to have the printed message, because
the single runner doesn't in quick mode.
The test.json format currently has three keys related to skipping tests:
* `skip_on_jobname`
* `skip_on_os`
* `skip_on_env`
While `skip_on_env` marks the test itself as skipped, i.e. they don't get run when the
conditions are met, the other two skip options are just marking the test as "expected to be skipped"
if the conditions apply, i.e. they want to see `MESON_SKIP_TEST` in the output and things will
fail if that doesn't happen. They don't actually skip the tests as the names imply.
To make this clearer rename the keys:
* `skip_on_jobname` -> `expect_skip_on_jobname`
* `skip_on_os` -> `expect_skip_on_os`
`skip_on_env` stays the same, since that actually skips.
The docs were also confused about this, so adjust that too.
On Debian systems the cython compiler binary is installed as
`cython3`. The current logic for detecting whether to run the Cython
unit tests instead checks for `cython` or the value of the `CYTHON`
environment variable, which leads to cases where the underlying Meson
can correctly compile Cython code but the test harness excludes these
tests from execution because it cannot find `cython3`.
This commit makes the test harness use the same detection method as
Meson itself. It also takes the opportunity to refactor some existing
code which does the same job for Objective C and Objective C++ tests
to avoid repetition.
git worktrees are frequently a handy tool to use. This glob pattern
asserts an error if the same named directory appears more than once,
which it will -- once per worktree -- but only because it is globbing
the entire project root, even though it should always only exclusively
care about files in `test cases/`. So just glob correctly, and then I
can run the testsuite locally again.
Fixes crash when running the tests and yasm is installed, but nasm is
not. When printing the compilers at early startup in
run_project_tests.py, we try to detect the nasm compiler, which in this
case turns out to be yasm, and its exelist assumes that the meson
command is set.
Fixes#12662
This replaces all of the Apache blurbs at the start of each file with an
`# SPDX-License-Identifier: Apache-2.0` string. It also fixes existing
uses to be consistent in capitalization, and to be placed above any
copyright notices.
This removes nearly 3000 lines of boilerplate from the project (only
python files), which no developer cares to look at.
SPDX is in common use, particularly in the Linux kernel, and is the
recommended format for Meson's own `project(license: )` field
- On Windows, it was not detected if include directory was an absolute
path to source directory, because of the mis of path separators.
- In the edgecase the include directory begins with the exact same
string as the source directory, but is a different directory, it was
falsely reported as an error.
Fixes#12217.
Installing python sources causes the python module to call
create_install_data() before Ninja backends adds extra outputs to Vala
targets.
Target objects are supposed to be immutable, adding outputs that late is
totally wrong. Add extra vala outputs immediately, but be careful
because the main output is only added later in post_init(). Luckily
the base class already puts a placeholder item in self.outputs for the
main filename so we can just replace self.outputs[0] instead of
replacing the whole list which would contain vala outputs at that stage.
This is surprisingly what SharedLibrary was already doing.
This commit adds a new keyword arg to extension_module() that enables
a user to target the Python Limited API, declaring the version of the
limited API that they wish to target.
Two new unittests have been added to test this functionality.
This lessens the amount of code imported at Meson startup by mapping
each dependency to a dictionary entry and using a programmable import to
dynamically return it.
Minus 16 files and 6399 lines of code imported at startup.
In the commit that originally added this import, it wasn't even used.
Now the implementation is being moved, so it will fail to work.
I do not know why I originally added it, but it needs to go. :)
There is no need to state for every single test that "preconditions were
not met". And logging the skip reason for a single test is easy to read,
but making every second line alternate is less so.
On Windows, in Python version prior to 3.8.7, the `sysconfig` modules
provides an extension filename suffix that disagrees the one returned
by `distutils.sysconfig`. Get the more awesome suffix from the latter
when building for a Python version known to present this issue.
Simplify the extension module filename suffix lookup to use the same
method used by `setuptools`.
Adjust project tests accordingly.
Fixes#10960.
It was impossible to specify null arguments in the exclude array,
while they were passed properly in the code, they were stringified
while the entries in the options array were just completely removed.
To fix this, do not early stringify the argument name and option
together as that would make it impossible to properly match it to
None without introducing a new special case for the string "None".
So instead they are now kept as tuples and only stringified when
its actually needed.
Andoni Morales Alastruey
Jussi Pakkanen
Eli Schwartz
H. Vetinari
Dylan Baker
Charles Brunet
Christoph Reiter
Andrew McNulty
Tristan Partin
Arsen Arsenović
Xavier Claessens
Eli Schwartz
Nomura
Josh Soref
Daniele Nicolodi
Marvin Scholz
Robert Cohn