How would I go about running a

python_source

and collecting its output? I have a python script that runs my app and writes the generated api spec (to stdout, but could easily switch it to a file). I'm trying to figure out how to run this and then use the output file for something (take a diff against the PR target, so as to check for changes to the spec). It seems I need to use something with

run_shell_command

, but I can't seem to figure it out (aside from exporting the venv, then having the shell command use

./dist/export/<.......>/bin/python

to run the script, which I would like to avoid because it feels like their ought to be a cleaner way)

You would likely need to use

adhoc_tool

.

Fleshing out that suggestion: something along the lines of

adhoc_tool(name="generated", runnable="./script.py", stdout="file_name_to_capture_to.txt")

is a likely starting point. Depending on

path/to/directory:generated

will then act like depending on a

file(name="generated", source="file_name_to_capture_to.txt")

if there was a file by that name on disk (i.e. it'll appear in sandboxes). https://github.com/pantsbuild/pants/discussions/18235#discussioncomment-6655594 might also be relevant to you.

Thanks for the great suggestions! I ended up just using

pants run /path/to/script.py | tail -n 1 > api_spec.json

since I wanted it to appear directly in my repo (eg. to save as a github artifact afterwards)