rclone/bin/check_autogenerated_edits.py

#!/usr/bin/env python3
"""
This script checks for unauthorized modifications in autogenerated sections of markdown files.
It is designed to be used in a GitHub Actions workflow or a local pre-commit hook.

Features:
- Detects markdown files changed between a commit and one of its ancestors. Default is to
  check the last commit only. When triggered on a pull request it should typically compare the
  pull request branch head and its merge base - the commit on the main branch before it diverged.
- Identifies modified autogenerated sections marked by specific comments.
- Reports violations using GitHub Actions error messages.
- Exits with a nonzero status code if unauthorized changes are found.
"""

import re
import subprocess
import sys

def run_git(args):
    """
    Run a Git command with the provided arguments and return its output as a string.
    """
    return subprocess.run(["git"] + args, stdout=subprocess.PIPE, text=True, check=True).stdout.strip()

def get_changed_files(base, head):
    """
    Retrieve a list of markdown files that were changed between the base and head commits.
    """
    files = run_git(["diff", "--name-only", f"{base}...{head}"]).splitlines()
    return [f for f in files if f.endswith(".md")]

def get_diff(file, base, head):
    """
    Get the diff of a given file between the base and head commits.
    """
    return run_git(["diff", "-U0", f"{base}...{head}", "--", file]).splitlines()

def get_file_content(ref, file):
    """
    Retrieve the content of a file from a given Git reference.
    """
    try:
        return run_git(["show", f"{ref}:{file}"]).splitlines()
    except Exception:
        return []

def find_regions(lines):
    """
    Identify the start and end line numbers of autogenerated regions in a file.
    """
    regions = []
    start = None
    for i, line in enumerate(lines, 1):
        if line.lstrip().startswith("<!-- autogenerated options start "):
            start = i
        elif start is not None and line.rstrip().endswith(" autogenerated options stop -->"):
            regions.append((start, i))
            start = None
    return regions

def in_region(ln, regions):
    """
    Check if a given line number falls within an autogenerated region.
    """
    return any(start <= ln <= end for start, end in regions)

def show_error(file_name, line, message):
    """
    Print an error message in a GitHub Actions-compatible format.
    """
    print(f"::error file={file_name},line={line}::{message} at {file_name} line {line}")

def check_file(file, base, head):
    """
    Check a markdown file for modifications in autogenerated regions.
    """
    viol = False
    new_lines = get_file_content("HEAD", file)
    old_lines = get_file_content("HEAD~1", file)

    # If old file did not exist or was empty then don't check
    if not old_lines:
        return
    
    # Entire autogenerated file check.
    if any("autogenerated - DO NOT EDIT" in l for l in new_lines[:10]):
        if get_diff(file, base, head):
            show_error(file, 1, "Autogenerated file modified")
            return True
        return False
    
    # Partial autogenerated regions.
    regions_new = find_regions(new_lines)
    regions_old = find_regions(old_lines)
    diff = get_diff(file, base, head)
    hunk_re = re.compile(r"^@@ -(\d+),?(\d*) \+(\d+),?(\d*) @@")
    new_ln = old_ln = None
    
    for line in diff:
        if line.startswith("@@"):
            m = hunk_re.match(line)
            if m:
                old_ln = int(m.group(1))
                new_ln = int(m.group(3))
        elif new_ln is None:
            continue
        elif line.startswith("+"):
            if in_region(new_ln, regions_new):
                show_error(file, new_ln, "Autogenerated region of file modified")
                viol = True
            new_ln += 1
        elif line.startswith("-"):
            if in_region(old_ln, regions_old):
                show_error(file, old_ln, "Autogenerated region of file modified")
                viol = True
            old_ln += 1
        else:
            new_ln += 1
            old_ln += 1
    
    return viol

def main():
    """
    Main function that iterates over changed files and checks them for violations.
    """
    base = "HEAD~1"
    head = "HEAD"
    if len(sys.argv) > 1:
        base = sys.argv[1]
        if len(sys.argv) > 2:
            head = sys.argv[2]
    found = False
    for f in get_changed_files(base, head):
        if check_file(f, base, head):
            found = True
    if found:
        sys.exit(1)
    print("No unauthorized edits found in autogenerated sections.")
    sys.exit(0)

if __name__ == "__main__":
    main()