all-repos autofixer for git.io

git_io_fix.py

from __future__ import annotations

import argparse
import functools
import re
import urllib.request
from typing import Match
from typing import Sequence

from all_repos import autofix_lib
from all_repos.config import Config

GIT_IO_RE = re.compile(rb'https://git\.io/.+?\b')


def find_repos(config: Config) -> set[str]:
    raise NotImplementedError('need to specify --repos')


@functools.lru_cache
def _git_io_replacement(url: bytes) -> bytes:
    req = urllib.request.Request(url.decode(), method='HEAD')
    resp = urllib.request.urlopen(req)
    return resp.url.encode()


def _replace_cb(match: Match[bytes]) -> str:
    return _git_io_replacement(match[0])


def apply_fix() -> None:
    ret = autofix_lib.run(
        'git', 'grep', '-El', GIT_IO_RE.pattern.decode(),
        capture_output=True,
        text=True,
    )
    for line in ret.stdout.splitlines():
        with open(line, 'rb') as f:
            contents = f.read()
        new_contents = GIT_IO_RE.sub(_replace_cb, contents)
        if new_contents != contents:
            with open(line, 'wb') as f:
                f.write(new_contents)


def main(argv: Sequence[str] | None = None) -> int:
    parser = argparse.ArgumentParser()
    autofix_lib.add_fixer_args(parser)
    args = parser.parse_args(argv)

    msg = '''\
replace git.io links with redirect targets

see: https://github.blog/changelog/2022-04-25-git-io-deprecation/
'''

    repos, config, commit, autofix_settings = autofix_lib.from_cli(
        args,
        find_repos=find_repos,
        msg=msg,
        branch_name='git-io-deprecated',
    )

    autofix_lib.fix(
        repos,
        apply_fix=apply_fix,
        config=config,
        commit=commit,
        autofix_settings=autofix_settings,
    )
    return 0


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

an all-repos autofixer for git.io urls

usage is something like:

python -m git_io_fix --repos $(all-repos-grep --repos -E 'https://git\.io/.+?\b')

an example pull request: getsentry/develop#565

Note that a few years ago, git.io was accessible in http://, I have a few http://git.io/ links in my codebase.