This currently does not work because we also need to account for the fact that R_HOME might not be set (as it seems in the test suite). In that case,bin/Rscript Is returned.

To be true to the name of the function return value, maybe we should (Edit: rename the function) and look up if Rscript is on the PATH and return that full path or - if Rscript is not on the PATH - return $R_HOME/bin/Rscript, and error if $R_HOME is not set.

Of course! Sorry I missed that. I'll put together something more thorough along those lines.

Cool. Let me know if you need help. Also, we should rename the function _rscript_exec to _path_rscript_exec() as it now always returns a path to the executable and it would be great if you could also test to tests/languages/r_test.py. You can ensure the same env as in the execution is present and set R home in a similar way:

from pre_commit.languages.r import in_env, _path_rscript_exec
from pre_commit.envcontext import envcontext

@pytest.mark.parametrize(('r_home_env', 'rscript_path_expected') [("/usr/bin/local", "/user/bin/local/Rscript"), (UNSET, 'location is os specific: get it dynamically)])
def test_path_rscript_exec(r_home_env, rscript_path_expected):
    with in_env(hook.prefix, hook.language_version):
        with envcontext(("R_HOME", r_home_env)): 
            assert _path_rscript_exec() == rscript_path_expected

If it gives an error or takes too long to create a test, I can figure that out later.

Actually we should not really need with in_env…

I went ahead and added that test and adjusted the code

thank you, seems to work.