#!/usr/bin/env python3
"""
清理过期数据脚本
删除已过期超过指定天数的数据
"""

import json
import os
from datetime import datetime, timedelta
from pathlib import Path


DATA_DIR = Path("/home/hermes/projects/data-sync-service/data")
DELETE_AFTER_DAYS = 30  # 过期后保留天数


def parse_datetime(dt_str):
    """解析时间字符串"""
    try:
        return datetime.strptime(dt_str, "%Y-%m-%d %H:%M:%S")
    except:
        return None


def clean_expired():
    """清理过期数据"""
    
    index_file = DATA_DIR / "index.json"
    
    if not index_file.exists():
        print("⚠️ 索引文件不存在")
        return
    
    with open(index_file, 'r', encoding='utf-8') as f:
        index = json.load(f)
    
    now = datetime.now()
    threshold = now - timedelta(days=DELETE_AFTER_DAYS)
    
    deleted_count = 0
    new_tags = []
    
    for item in index.get("tags", []):
        tag = item.get("tag")
        expires_at_str = item.get("expires_at")
        filename = item.get("file")
        
        if not expires_at_str:
            # 无过期时间，保留
            new_tags.append(item)
            continue
        
        expires_at = parse_datetime(expires_at_str)
        if not expires_at:
            # 解析失败，保留
            new_tags.append(item)
            continue
        
        # 检查是否过期超过阈值
        if expires_at < threshold:
            print(f"🗑️ 删除过期数据: {tag} (过期于 {expires_at_str})")
            
            # 删除数据文件
            filepath = DATA_DIR / filename
            if filepath.exists():
                os.remove(filepath)
            
            deleted_count += 1
        else:
            new_tags.append(item)
    
    # 更新索引
    index["tags"] = new_tags
    with open(index_file, 'w', encoding='utf-8') as f:
        json.dump(index, f, indent=2, ensure_ascii=False)
    
    print(f"\n✅ 清理完成: 删除 {deleted_count} 个过期数据")


if __name__ == "__main__":
    print("=" * 60)
    print(f"清理过期数据 (过期超过 {DELETE_AFTER_DAYS} 天)")
    print("=" * 60)
    
    clean_expired()