scripts/7_prepare_circle_names.py

import os
import sys
import json
from collections import defaultdict
from typing import Dict, List, Set

sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))

from utils.logger import setup_logger
from utils.json import get_file_path_from_config, json_dumps
from schemas.circles import Circle

log = setup_logger(__name__)


def normalized_circle_name(name: str) -> str:
    """
    サークル名の正規化を行う。
    - 漢字・ひらがな・カタカナはそのまま
    - 半角英数字は小文字に変換
    - 記号類（！"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~・／？＜＞（）【】？｜：）を除く
    """
    SYMBOLS = set("!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~・／？＜＞（）【】？｜：")
    if not name:
        return ""
    normalized = []
    for char in name:
        if "A" <= char <= "Z":
            normalized.append(char.lower())
        elif char in SYMBOLS:
            continue
        else:
            normalized.append(char)
    return "".join(normalized)


def get_substrings(texts: Dict[any, str]) -> Set[str]:
    """テキストから2文字以上の連続部分文字列をすべて抽出する。"""
    substrings = set()
    for key, text in texts.items():
        if key == "circleId":
            continue
        text = text.replace(" ", "").replace("　", "")
        length = len(text)
        for start in range(length):
            for end in range(start + 2, length + 1):
                substr = text[start:end]
                substrings.add(substr)
    return substrings


def main():
    input_file = get_file_path_from_config("circles.circles_json")

    try:
        with open(input_file, encoding="utf-8") as f:
            circles = json.load(f)
    except FileNotFoundError:
        log.error(f"入力ファイルが見つかりません: {input_file}")
        sys.exit(1)
    except json.JSONDecodeError as e:
        log.error(f"JSONデコードエラー: {e}")
        sys.exit(1)

    # --- 団体名データを生成 ---
    output_file_names = get_file_path_from_config("substring.circle_names")
    circles = [Circle(**item) for item in circles]
    circle_names = [
        {
            "circleId": c.circleId,
            "circle": c.circleName,
            "circleNormalized": normalized_circle_name(c.circleName),
            "circleKana": c.circleNameKana or "",
        }
        for c in circles
    ]

    # 出力先ディレクトリ作成（念のため）
    os.makedirs(os.path.dirname(output_file_names), exist_ok=True)

    # JSON に書き出し
    json_dumps(circle_names, output_file_names)
    log.info(f"サークル名データを出力しました: {output_file_names}")

    # --- 部分文字列インデックスを生成 ---
    output_file_substring = get_file_path_from_config("substring.substring_index")
    substring_to_circle_ids: Dict[str, List[str]] = defaultdict(list)
    for circle in circle_names:
        circleId = circle["circleId"]
        substrings = get_substrings(circle)
        for substr in substrings:
            substring_to_circle_ids[substr].append(circleId)

    # 出力先ディレクトリ作成（念のため）
    os.makedirs(os.path.dirname(output_file_substring), exist_ok=True)

    # JSON に書き出し
    json_dumps(substring_to_circle_ids, output_file_substring)
    log.info(f"部分文字列インデックスを出力しました: {output_file_substring}")


if __name__ == "__main__":
    main()