Function main

scripts/fetch_fulltext.py:337–388

()

def main():
    parser = argparse.ArgumentParser(description="Fetch fulltext for knowledge base documents")
    parser.add_argument("--framework", "-f", default="", help="Filter by framework (fda, eu_mdr, nmpa)")
    parser.add_argument("--type", "-t", default="", help="Filter by document type (guidance, regulations)")
    parser.add_argument("--id", default="", help="Process a specific document by ID")
    parser.add_argument("--all", action="store_true", help="Process all documents")
    parser.add_argument("--dry-run", action="store_true", help="Show what would be done without fetching")
    parser.add_argument("--force", action="store_true", help="Re-fetch even if fulltext exists")
    args = parser.parse_args()

    if not args.framework and not args.id and not args.all:
        parser.print_help()
        print("\nExamples:")
        print("  python scripts/fetch_fulltext.py --framework fda --type guidance")
        print("  python scripts/fetch_fulltext.py --id fda-cybersecurity-premarket-2026")
        print("  python scripts/fetch_fulltext.py --all --dry-run")
        return

    entries = load_entries(
        framework=args.framework,
        doc_type=args.type,
        doc_id=args.id,
    )

    if not entries:
        logger.info("No entries found matching criteria.")
        return

    logger.info(f"Found {len(entries)} entries to process")

    success = 0
    skipped = 0
    failed = 0

    for entry in entries:
        if not args.force and fulltext_exists(entry):
            skipped += 1
            continue

        try:
            ok = process_entry(entry, dry_run=args.dry_run)
            if ok:
                success += 1
            else:
                skipped += 1
        except Exception as e:
            logger.error(f"  EXCEPTION: {e}")
            failed += 1

        time.sleep(1)

    logger.info(f"\nSummary: {success} processed, {skipped} skipped, {failed} failed (total: {len(entries)})")


if __name__ == "__main__":