#!/usr/bin/env python3
"""Refuse a file too long to review, and say so before one gets there.

Two numbers, both in lines the way `wc -l` counts them, and both set once at
the top of this file. A file over the block limit fails the run: nobody reads
one of those end to end, and a change inside it is reviewed by hope. A file
over the warning limit is printed and the run still passes: it is the point at
which a file has picked up a second subject and the next person to touch it is
the one who can still name what those subjects are.

Neither number judges the code. A long file of one thing is a long file of one
thing, and the warning is a question, not a finding. What the block limit says
is that the answer has stopped being anybody's to give.

No third-party modules on purpose: this runs on every commit, so it has to
start fast and work in a fresh clone.

Run by prek over the staged files, or by hand:
    scripts/check-file-size.py <path>...
"""

import sys

BLOCK_LINES = 5000
WARN_LINES = 2000


def count_lines(path):
    """Lines in a file, or None if it cannot be read as one."""
    try:
        with open(path, "rb") as handle:
            return handle.read().count(b"\n")
    except OSError:
        return None


def main(paths):
    blocked = []
    for path in paths:
        lines = count_lines(path)
        if lines is None:
            continue
        if lines > BLOCK_LINES:
            blocked.append((path, lines))
        elif lines > WARN_LINES:
            print(
                f"check-file-size: {path} is {lines} lines, over {WARN_LINES}."
                " Worth splitting while it is still one person's to split."
            )
    for path, lines in blocked:
        print(
            f"check-file-size: {path} is {lines} lines, over the"
            f" {BLOCK_LINES} a commit may carry. Split it first.",
            file=sys.stderr,
        )
    return 1 if blocked else 0


if __name__ == "__main__":
    sys.exit(main(sys.argv[1:]))