scripts/00_import_csv.py

"""
Step 0: menu.csv → data/raw/real_menus.json 변환
실행: python scripts/00_import_csv.py

필터 조건:
    - IS_VISIBLE_MENU == "true"
    - DEPTH1_NAME 존재
    - DEPTH4_NAME에 "팝업" 미포함

menu_id = SCR_{SCREEN_NUM}
menu_path: DEPTH2 == DEPTH1이면 중복 레벨 제거
"""
import csv
import json
import sys
from pathlib import Path

sys.path.insert(0, str(Path(__file__).parent.parent))

from config import RAW_DIR

CSV_PATH = Path(__file__).parent.parent.parent / "menu.csv"
OUTPUT_PATH = RAW_DIR / "real_menus.json"


def build_menu_path(row: dict) -> str:
    """DEPTH 계층에서 menu_path 생성. DEPTH2 == DEPTH1이면 DEPTH2 제거."""
    d1 = (row.get("DEPTH1_NAME") or "").strip()
    d2 = (row.get("DEPTH2_NAME") or "").strip()
    d3 = (row.get("DEPTH3_NAME") or "").strip()
    d4 = (row.get("DEPTH4_NAME") or "").strip()

    # DEPTH2가 DEPTH1과 동일하면 중복 레벨 제거
    if d2 == d1:
        d2 = ""

    parts = [p for p in [d1, d2, d3, d4] if p]
    return " > ".join(parts)


def parse_keywords(raw: str) -> list:
    """MENU_KEYWORD 콤마 구분 파싱 → List[str]"""
    if not raw or not raw.strip():
        return []
    return [kw.strip() for kw in raw.split(",") if kw.strip()]


def parse_search_count(raw: str) -> int:
    """SEARCH_COUNT 정수 변환 (없으면 0)"""
    try:
        return int(raw.strip()) if raw and raw.strip() else 0
    except ValueError:
        return 0


def main():
    if not CSV_PATH.exists():
        print(f"[오류] menu.csv 파일을 찾을 수 없습니다: {CSV_PATH}")
        print("menu.csv를 프로젝트 루트(prototype 상위 폴더)에 위치시켜 주세요.")
        sys.exit(1)

    RAW_DIR.mkdir(parents=True, exist_ok=True)

    menus = []
    skipped = {"invisible": 0, "no_depth1": 0, "popup": 0}
    screen_num_set = set()

    with open(CSV_PATH, encoding="utf-8-sig", newline="") as f:
        reader = csv.DictReader(f)
        for row in reader:
            # 필터 1: IS_VISIBLE_MENU == "true"
            if row.get("IS_VISIBLE_MENU", "").strip().lower() != "true":
                skipped["invisible"] += 1
                continue

            # 필터 2: DEPTH1_NAME 존재
            d1 = (row.get("DEPTH1_NAME") or "").strip()
            if not d1:
                skipped["no_depth1"] += 1
                continue

            # 필터 3: 팝업 제외
            d4 = (row.get("DEPTH4_NAME") or "").strip()
            if "팝업" in d4:
                skipped["popup"] += 1
                continue

            screen_num = (row.get("SCREEN_NUM") or "").strip()
            if not screen_num:
                continue

            menu_id = f"SCR_{screen_num}"
            menu_name = (row.get("DEPTH4_NAME") or row.get("DEPTH3_NAME") or d1).strip()
            menu_path = build_menu_path(row)
            category = d1
            keywords = parse_keywords(row.get("MENU_KEYWORD", ""))
            search_count = parse_search_count(row.get("SEARCH_COUNT", ""))

            # SCREEN_NUM 중복 체크 (이론상 없어야 하지만 안전 처리)
            if screen_num in screen_num_set:
                print(f"  [경고] SCREEN_NUM 중복: {screen_num} ({menu_name}) 스킵")
                continue
            screen_num_set.add(screen_num)

            menus.append({
                "menu_id":      menu_id,
                "menu_name":    menu_name,
                "menu_path":    menu_path,
                "category":     category,
                "screen_num":   screen_num,
                "search_count": search_count,
                "keywords":     keywords,
            })

    OUTPUT_PATH.write_text(
        json.dumps(menus, ensure_ascii=False, indent=2),
        encoding="utf-8"
    )

    print(f"변환 완료: {len(menus)}개 메뉴")
    print(f"  스킵 - IS_VISIBLE_MENU 제외: {skipped['invisible']}개")
    print(f"  스킵 - DEPTH1 없음: {skipped['no_depth1']}개")
    print(f"  스킵 - 팝업 제외: {skipped['popup']}개")
    print(f"저장: {OUTPUT_PATH}")

    # 카테고리별 집계
    from collections import Counter
    cat_counts = Counter(m["category"] for m in menus)
    print("\n카테고리별 메뉴 수:")
    for cat, cnt in cat_counts.most_common():
        print(f"  {cat}: {cnt}개")

    # keywords 보유율
    has_kw = sum(1 for m in menus if m["keywords"])
    print(f"\nkeywords 보유: {has_kw}/{len(menus)} ({has_kw/len(menus)*100:.1f}%)")

    # search_count 보유율
    has_sc = sum(1 for m in menus if m["search_count"] > 0)
    print(f"search_count > 0: {has_sc}/{len(menus)} ({has_sc/len(menus)*100:.1f}%)")


if __name__ == "__main__":
    main()