tests: improve the test suite implementation

[perillo]

This is the first of a series of PR to improve the test suite readability and reliability.

• [x] tests: ensure python 3.8

  This version supports features that will help improving the current implementation:

  • subprocess.run
  • fstring
  • positional-only parameters

  Update tests.CMakeLists.txt to find the Python3 package.

  Add the ruff.toml file. Line lenght has been set to 100, and isort is enabled for lint.

  See also https://devguide.python.org/versions/

• [x] tests: run "ruff format tests/tools"

  Run "ruff format tests/tools" and "ruff format tests/tools/run-tests"

  This commit changes a lot of code

  statistics: tests/tools/accumulate.py | 166 +++++++++-- tests/tools/bash.py | 327 +++++++++++++++++++--- tests/tools/bash_linux_only.py | 42 ++- tests/tools/basic.py | 19 +- tests/tools/compiled.py | 463 +++++++++++++++++++++++++------ tests/tools/compiled_basic.py | 81 +++++- tests/tools/filter.py | 51 +++- tests/tools/parse_cobertura.py | 19 +- tests/tools/python.py | 127 ++++++++- tests/tools/run-tests | 6 +- tests/tools/system_mode.py | 32 ++- tests/tools/testbase.py | 23 +- 12 files changed, 1153 insertions(+), 203 deletions(-)

• [x] tests: run "ruff check --fix tests/tools"

  Run "ruff check --fix tests/tools" and "ruff check --fix tests/tools/run-tests"

  Apply safe fixes.

• [x] tests: run "ruff check --fix --unsafe-fixes tests/tools"

  Run "ruff check --fix --unsafe-fixes tests/tools" and "ruff check --fix --unsafe-fixes tests/tools/run-tests"

  Apply unsafe fixes. These are actually safe.

  These are additional errors reported by ruff:

  tests/tools/compiled.py:217:9: F821 Undefined name self tests/tools/compiled.py:321:9: E722 Do not use bare except tests/tools/compiled.py:685:9: F821 Undefined name time tests/tools/compiled.py:689:9: F821 Undefined name time tests/tools/system_mode.py:61:9: E722 Do not use bare except

Most of the code change has be done automatically. Also note that some change are not optimal, but I avoided manual changes.

Since review will be very hard and you should not trust me, I suggest to @SimonKagstrom that

Only merge the first commit. Add a commit for each of the following commands:

> ruff format tests/tools
> ruff format tests/tools/run-tests

> ruff check --fix tests/tools
> ruff check --fix tests/tools/run-tests

> ruff format tests/tools
> ruff format tests/tools/run-tests

I used ruff 0.3.3 but it should be ok to use a different version.

Formatting the code will make future changes more easy. As an example, I plan to change how the command line is created for subprocess. Currently is is a string, where arguments are concatenated together; it is terrible!

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 65.68%. Comparing base (49635e7) to head (2d48999).

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #429 +/- ## ======================================= Coverage 65.68% 65.68% ======================================= Files 58 58 Lines 4514 4514 Branches 4171 4171 ======================================= Hits 2965 2965 Misses 1549 1549 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[SimonKagstrom]

I'm OK with merging all of it at once, it's still separated into commits, or did I misunderstand you?

Thanks for doing this!

[perillo]

I'm OK with merging all of it at once, it's still separated into commits, or did I misunderstand you?

Thanks for doing this!

What I mean is that this is a big change. Personally I would feel insecure about merging a pull request like this. Thanks for the trust!

[SimonKagstrom]

Well, I looked at the commits, and since it only affects the testsuite (which also runs fine), I'm not too worried!