Function main

scripts/developer-guide/run_languagetool.py:542–607

()

def main():
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--html", required=True, help="Rendered HTML input")
    parser.add_argument("--output", required=True, help="JSON output path")
    parser.add_argument("--language", default="en-US")
    parser.add_argument(
        "--accept-list",
        default=os.path.join(
            os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))),
            "docs",
            "developer-guide",
            "languagetool-accept.txt",
        ),
        help="Path to the LanguageTool accept-list file (one regex per line).",
    )
    args = parser.parse_args()

    text = extract_text(args.html)
    accept_re = load_accept_patterns(args.accept_list)

    report = {"status": "unknown", "matches": [], "total": 0}
    try:
        try:
            matches = run_languagetool(text, language=args.language, accept_re=accept_re)
        except Exception as exc:  # noqa: BLE001 — advisory check must not crash CI
            print(f"LanguageTool failed to start: {exc}", file=sys.stderr)
            report = {"status": "error", "reason": str(exc), "matches": [], "total": 0}
        else:
            if matches is None:
                report = {
                    "status": "skipped",
                    "reason": "language_tool_python not installed",
                    "matches": [],
                    "total": 0,
                }
            else:
                try:
                    serialized = matches_to_json(matches, text)
                except Exception as exc:  # noqa: BLE001
                    print(
                        f"LanguageTool: failed to serialize {len(matches)} match(es): {exc}",
                        file=sys.stderr,
                    )
                    report = {
                        "status": "error",
                        "reason": f"serialization failed: {exc}",
                        "matches": [],
                        "total": len(matches),
                    }
                else:
                    report = {"status": "ok", "matches": serialized, "total": len(serialized)}
    finally:
        os.makedirs(os.path.dirname(args.output) or ".", exist_ok=True)
        with open(args.output, "w", encoding="utf-8") as fh:
            json.dump(report, fh, indent=2)
        print(
            f"LanguageTool report written to {args.output} "
            f"({report.get('total', 0)} match(es), status={report.get('status')})."