scripts/2_create_circles_data.py

#!/usr/bin/env python3
"""
CSVからサークルデータのJSONを生成するスクリプト
"""

import os
import sys
import csv
import json

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

# --- ロギングの設定 ---
log = setup_logger(__name__)


def parse_bool(value: str) -> bool:
    """文字列をブール値に変換"""
    return value.lower() in ("true", "1", "yes")


def parse_int(value: str) -> int | None:
    """文字列を整数に変換（空の場合はNone）"""
    if not value or value.strip() == "":
        return None
    try:
        return int(value)
    except ValueError:
        return None


def parse_list(value: str, delimiter: str = "|") -> list[str]:
    """区切り文字で分割してリストに変換"""
    if not value or value.strip() == "":
        return []
    return [item.strip() for item in value.split(delimiter) if item.strip()]


def parse_json(value: str) -> list | dict | None:
    """JSON文字列をパースする（空の場合はNone、失敗した場合は空リスト）"""
    if not value or value.strip() == "":
        return None
    try:
        return json.loads(value)
    except json.JSONDecodeError:
        log.warning(f"JSONのパースに失敗しました: {value[:50]}...")
        return []


def convert_annual_schedule(old_schedule: list | None) -> list | None:
    """
    旧形式の年間予定を新形式に変換する。

    旧形式: [{"month": 4, "events": ["新歓", "春季リーグ開幕"]}, ...]
    新形式: [{"period": "4月", "event": "新歓"}, {"period": "4月", "event": "春季リーグ開幕"}, ...]
    """
    if not old_schedule:
        return None

    new_schedule = []
    for item in old_schedule:
        # 旧形式（month + events）の場合
        if "month" in item and "events" in item:
            month = item["month"]
            events = item["events"]
            period = f"{month}月"
            for event in events:
                new_schedule.append({"period": period, "event": event})
        # 新形式（period + event）の場合はそのまま
        elif "period" in item and "event" in item:
            new_schedule.append(item)

    return new_schedule if new_schedule else None


def main():
    """CSVからサークルデータを読み込み circles.json を生成する。"""

    log.info("CSVファイルからサークルデータを読み込みcircles.jsonを生成します。")

    input_csv_path = get_file_path_from_config(
        "circles.original_csv", "resources/original_circles.csv"
    )
    output_json_path = get_file_path_from_config(
        "circles.circles_json", "data/generated/circles.json"
    )

    if not os.path.exists(input_csv_path):
        log.error(f"入力CSVファイルが見つかりません: {input_csv_path}")
        sys.exit(1)

    circles = []
    try:
        with open(input_csv_path, "r", encoding="utf-8") as f:
            reader = csv.DictReader(f)
            for row in reader:
                # 画像データのパース
                images_data = parse_json(row.get("images", ""))
                images = images_data if images_data else []

                # mainImage の決定: images配列の最初の画像（order=0）を使用
                main_image = None
                if images:
                    # order=0 の画像を探す
                    for img in images:
                        if img.get("order", 0) == 0:
                            main_image = img
                            break
                    # order=0 がなければ最初の画像を使用
                    if main_image is None:
                        main_image = images[0]

                # SNSデータのパース（handleフィールドを含む）
                sns_data = parse_json(row.get("sns", ""))
                sns = []
                if sns_data:
                    for sns_item in sns_data:
                        sns.append({
                            "type": sns_item.get("type", ""),
                            "url": sns_item.get("url", ""),
                            "handle": sns_item.get("handle")
                        })

                # 年間予定のパース（新形式に変換）
                annual_schedule_data = parse_json(row.get("annualSchedule", ""))
                annual_schedule = convert_annual_schedule(annual_schedule_data)

                circle = {
                    "circleId": row.get("circleId", ""),
                    "circleName": row.get("circleName", ""),
                    "circleNameKana": row.get("circleNameKana", ""),
                    "projectId": row.get("projectId") or None,
                    "projectName": row.get("projectName") or None,
                    "genre": row.get("genre") or None,
                    "areaCode": row.get("areaCode") or None,
                    "pamphletNumber": parse_int(row.get("pamphletNumber", "")),
                    "shortIntro": row.get("shortIntro") or None,
                    "detailDescription": row.get("detailDescription") or None,
                    "memberCount": parse_int(row.get("memberCount", "")),
                    "memberNote": row.get("memberNote") or None,
                    "activityFrequency": row.get("activityFrequency") or None,
                    "activityFrequencyNote": row.get("activityFrequencyNote") or None,
                    "activityLocation": row.get("activityLocation") or None,
                    "annualFee": parse_int(row.get("annualFee", "")),
                    "otherCosts": row.get("otherCosts") or None,
                    "annualSchedule": annual_schedule,
                    "contactInfo": row.get("contactInfo") or None,
                    "websiteUrl": row.get("websiteUrl") or None,
                    "isArchived": parse_bool(row.get("isArchived", "false")),
                    "images": images,
                    "mainImage": main_image,
                    "sns": sns,
                }
                circles.append(circle)
    except Exception as exc:
        log.error(f"CSVファイルの読み込みに失敗しました: {exc}")
        sys.exit(1)

    os.makedirs(os.path.dirname(output_json_path), exist_ok=True)
    json_dumps(circles, output_json_path)

    log.info(f"サークルデータの生成が完了しました: {output_json_path} ({len(circles)}件)")
    sys.exit(0)


if __name__ == "__main__":
    main()