chore: use subprocess.run instead of old check_*

[henryiii]

I noticed this was using the old subprocess.check_* methods. In modern Python, subprocess.run is recommended.

[henryiii]

This removes the ability to check and see why a command failed! My version:

>>> try:
...     subprocess.run(["ruff", "invalid"], stdout=subprocess.PIPE, stderr=subprocess.PIPE, check=True)
... except subprocess.CalledProcessError as err:
...     print(err.stdout, err.stderr)
...
b'' b"error: unrecognized subcommand 'invalid'\n\nUsage: ruff [OPTIONS] <COMMAND>\n\nFor more information, try '--help'.\n"

Current version:

>>> try:
...     subprocess.run(["ruff", "invalid"], stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL, check=True)
... except subprocess.CalledProcessError as err:
...     print(err.stdout, err.stderr)
...
None None

If a command fails, users will want to know why it failed!

This was available before (though with stdout and stderr merged, I didn't do that in my example here):

>>> try:
...     subprocess.check_output(["ruff", "invalid"], stderr=subprocess.STDOUT)
... except subprocess.CalledProcessError as err:
...     print(err.stdout, err.stderr)
...
b"error: unrecognized subcommand 'invalid'\n\nUsage: ruff [OPTIONS] <COMMAND>\n\nFor more information, try '--help'.\n" None

We catch and print this all the time in build: https://github.com/search?q=repo%3Apypa%2Fbuild%20CalledProcessError&type=code

[takluyver]

You're right, my mistake. I figured that if the return value wasn't used, there was no way to get the output, but I forgot about the exception having it. Please knock my commit off the branch. If you want to add a comment & a test, it might stop me making the same mistake in a couple of years time. 😉

…

On Wed, 18 Feb 2026, 18:33 Henry Schreiner, ***@***.***> wrote: *henryiii* left a comment (pypa/pyproject-hooks#210) <#210 (comment)> This removes the ability to check and see why a command failed! My version: >>> try: ... subprocess.run(["ruff", "invalid"], stdout=subprocess.PIPE, stderr=subprocess.PIPE, check=True) ... except subprocess.CalledProcessError as err: ... print(err.stdout, err.stderr) ... b'' b"error: unrecognized subcommand 'invalid'\n\nUsage: ruff [OPTIONS] <COMMAND>\n\nFor more information, try '--help'.\n" Current version: >>> try: ... subprocess.run(["ruff", "invalid"], stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL, check=True) ... except subprocess.CalledProcessError as err: ... print(err.stdout, err.stderr) ... None None If a command fails, users will want to know why it failed! This was available before (though with stdout and stderr merged): >>> try: ... subprocess.check_output(["ruff", "invalid"], stderr=subprocess.STDOUT) ... except subprocess.CalledProcessError as err: ... print(err.stdout, err.stderr) ... b"error: unrecognized subcommand 'invalid'\n\nUsage: ruff [OPTIONS] <COMMAND>\n\nFor more information, try '--help'.\n" None We catch and print this all the time in build: https://github.com/search?q=repo%3Apypa%2Fbuild%20CalledProcessError&type=code — Reply to this email directly, view it on GitHub <#210 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACQB5KQCK2T6HGPPOMTG434MSO65AVCNFSM6AAAAACVR5GY6KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTSMRSGE3DCOBYGE> . You are receiving this because you commented.Message ID: ***@***.***>

[takluyver]

Thanks!