Create-knowledge-find-decode/scripts/run_decode_content.py

"""Run the formal single-item decode workflow.

This CLI is intentionally thin. Production loading should come from the formal
candidate item repository; fixture loading is kept only for local smoke checks.
"""
from __future__ import annotations

import argparse
import json
from pathlib import Path
from uuid import UUID, uuid4

from acquisition.crawler import parse_detail_response
from core.config import Settings
from decode_content.service import DecodeService


ROOT = Path(__file__).resolve().parent.parent


def _load_fixture(path: Path):
    payload = json.loads(path.read_text(encoding="utf-8"))
    return parse_detail_response(payload, fallback_content_id=path.stem.removeprefix("xhs_case_"))


def main() -> None:
    parser = argparse.ArgumentParser(description="Decode one creation candidate through decode_content.")
    parser.add_argument("--fixture", type=Path, help="local fixture JSON for smoke runs")
    parser.add_argument("--item-id", help="candidate item UUID; generated when omitted for fixture smoke")
    args = parser.parse_args()
    if not args.fixture:
        parser.error("formal DB item loading is wired in pipeline Step 6; pass --fixture for local smoke")
    item_id = UUID(args.item_id) if args.item_id else uuid4()
    post = _load_fixture(args.fixture if args.fixture.is_absolute() else ROOT / args.fixture)
    service = DecodeService(settings=Settings.from_env())
    output = service.decode_post(item_id=item_id, post=post)
    print(json.dumps({"item_id": str(item_id), "status": output.status, "payload_count": len(output.payloads)}, ensure_ascii=False))


if __name__ == "__main__":
    main()