#!/usr/bin/env python3
from __future__ import annotations

import argparse
import concurrent.futures
import json
import os
import os.path
import subprocess
import sys

EXCLUDED = {
    ('windows-latest', 'r'),  # no r on windows
    ('windows-latest', 'swift'),  # no swift on windows
}

ALLOWED_LANGUAGES = {'python', 'pipenv'}

def _always_run() -> set[str]:
    return {
        'pre_commit/languages/all.py',
        'pre_commit/languages/helpers.py',
        'pre_commit/languages/__init__.py',
        'pre_commit/languages/python.py',  # python is a base for other languages
        'pre_commit/languages/pipenv.py',  # our new language
    }

def _lang_files(lang: str) -> set[str]:
    cmd = ('git', 'grep', '-l', '')
    env = dict(os.environ, GIT_LITERAL_PATHSPECS='1')
    out = subprocess.check_output(
        (*cmd, f':(glob,top)pre_commit/languages/{lang}.py'),
        env=env,
    ).decode()
    ret = set(out.splitlines())
    ret.add(f'tests/languages/{lang}_test.py')
    return ret

def main() -> int:
    parser = argparse.ArgumentParser()
    parser.add_argument('--all', action='store_true')
    args = parser.parse_args()

    langs = [
        lang for lang in ALLOWED_LANGUAGES
        if os.path.exists(f'pre_commit/languages/{lang}.py')
    ]

    triggers_all = _always_run()
    for fname in triggers_all:
        assert os.path.exists(fname), fname

    if not args.all:
        with concurrent.futures.ThreadPoolExecutor(os.cpu_count()) as exe:
            by_lang = {
                lang: files | triggers_all
                for lang, files in zip(langs, exe.map(_lang_files, langs))
            }

        diff_cmd = ('git', 'diff', '--name-only', 'origin/main...HEAD')
        files = set(subprocess.check_output(diff_cmd).decode().splitlines())

        langs = [
            lang
            for lang, lang_files in by_lang.items()
            if lang_files & files
        ]

    matched = [
        {'os': os, 'language': lang}
        for os in ('windows-latest', 'ubuntu-latest')
        for lang in langs
        if (os, lang) not in EXCLUDED
    ]

    with open(os.environ['GITHUB_OUTPUT'], 'a') as f:
        f.write(f'languages={json.dumps(matched)}\n')
    return 0

if __name__ == '__main__':
    raise SystemExit(main())