#!/usr/bin/env python3
"""SQLite 코퍼스의 CPU 영속 후보 인덱스를 신규 구축하거나 증분 동기화한다."""

from __future__ import annotations

import argparse
import json
import sys
from pathlib import Path

if __package__ in (None, ""):
    sys.path.insert(0, str(Path(__file__).resolve().parents[1]))

from app.engine.persistent_index import PersistentCorpusIndex
from app.engine.provenance import CorpusStore


def precompute_features(store: CorpusStore, batch: int = 500) -> int:
    """참조 lemma/요소를 미리 계산해 DB 에 저장 (#5).

    이 작업을 인덱싱 때 1회 해두면, 탐지 요청마다 후보 세그먼트를 형태소
    분석하던 비용이 사라진다. 이미 채워진 세그먼트는 건너뛴다.
    """
    from app.engine.extractor import get_extractor
    from app.engine.structural import extract_lemmas

    extractor = get_extractor()
    pending: list[tuple[str, list[str], dict]] = []
    updated = 0
    for segment in store.iter_segments():
        if segment.lemmas is not None and segment.elements is not None:
            continue
        pending.append((
            segment.segment_id,
            extract_lemmas(segment.text),
            extractor.extract(segment.text).model_dump(),
        ))
        if len(pending) >= batch:
            updated += store.update_segment_features(pending)
            print(f"  특징 계산 {updated}건…", flush=True)
            pending = []
    if pending:
        updated += store.update_segment_features(pending)
    return updated


def main() -> int:
    p = argparse.ArgumentParser(description=__doc__)
    p.add_argument("--database", type=Path, required=True)
    p.add_argument("--index-dir", type=Path, required=True)
    p.add_argument("--features", type=int, default=2**20)
    p.add_argument(
        "--skip-precompute", action="store_true",
        help="참조 lemma/요소 사전계산을 건너뛴다(질의 시 계산 후 백필됨)",
    )
    args = p.parse_args()
    if not args.database.exists():
        p.error(f"database does not exist: {args.database}")

    store = CorpusStore(args.database)
    precomputed = 0
    if not args.skip_precompute:
        precomputed = precompute_features(store)

    result = PersistentCorpusIndex(args.database, args.index_dir).sync(args.features)
    result["precomputed_features"] = precomputed
    result["missing_features"] = store.count_missing_features()
    print(json.dumps(result, ensure_ascii=False, indent=2))
    return 0


if __name__ == "__main__":
    raise SystemExit(main())