Search markdown files and update deprecated Wikimedia image sizes.

wikipedia-image-embeds-fix.py

#!/usr/bin/env -S uv run
# /// script
# requires-python = ">=3.13"
# dependencies = []
# ///
"""Search markdown files and update deprecated Wikimedia image sizes."""

import argparse
import re
import sys
from pathlib import Path

STANDARD_WIDTHS = [330, 500, 960, 1280, 1920, 3840]
WIKI_THUMB_RE = re.compile(
    r"(https://upload\.wikimedia\.org/wikipedia/commons/thumb/(?:[^/]+/){3})(\d+)(px-[^\s)\]\"]+)"
)


def process_file(file_path: Path, dry_run: bool) -> None:
    """Read file, calculate width updates, print changes, and optionally write."""
    content = file_path.read_text(encoding="utf-8")
    changes: list[tuple[int, int, str]] = []

    def snap_width(match: re.Match) -> str:
        base_url = match.group(1)
        current_width = int(match.group(2))
        suffix = match.group(3)

        nearest = min(STANDARD_WIDTHS, key=lambda x: abs(x - current_width))
        new_url = f"{base_url}{nearest}{suffix}"

        if current_width != nearest:
            changes.append((current_width, nearest, new_url))

        return new_url

    updated_content = WIKI_THUMB_RE.sub(snap_width, content)

    if changes:
        print(f"Updated {len(changes)} links in {file_path}")
        for old_w, new_w, new_url in changes:
            print(f"    {old_w}px to {new_w}px : {new_url}")

        if not dry_run:
            file_path.write_text(updated_content, encoding="utf-8")


def process_args(argv: list[str] | None = None) -> argparse.Namespace:
    """Parse command-line arguments."""
    parser = argparse.ArgumentParser(
        description="Fix Wikimedia thumbnail widths in markdown files."
    )
    parser.add_argument("directories", nargs="+", type=Path, help="Directories to scan")
    parser.add_argument(
        "--dry-run", action="store_true", help="Print changes without saving"
    )
    return parser.parse_args(argv)


def main() -> None:
    """Entry point."""
    args = process_args()
    for directory in args.directories:
        target_dir = directory.expanduser()
        if not target_dir.is_dir():
            print(f"Warning: {target_dir} is not a valid directory.", file=sys.stderr)
            continue

        for file_path in target_dir.rglob("*.md"):
            process_file(file_path, args.dry_run)


if __name__ == "__main__":
    main()