stage1st小黑屋

S1_blackroom.py

import requests
import json
import re
import openpyxl
import os
import time
from datetime import datetime
import glob  # 用于文件名模式匹配

# 检查并获取最新的文件
def get_latest_file():
    files = glob.glob('*_blackroom.xlsx')  # 搜索当前目录下所有匹配的文件
    if files:
        # 如果存在文件，返回最新的一个
        latest_file = max(files, key=os.path.getctime)
        print(f"Continuing from the latest file: {latest_file}")
        return latest_file
    else:
        # 如果没有匹配的文件，创建新文件
        timestamp = datetime.now().strftime("%Y%m%d%H%M%S")
        new_file = f"{timestamp}_blackroom.xlsx"
        print(f"Creating new file: {new_file}")
        return new_file

filename = get_latest_file()

# 初始化 Excel 工作簿
if os.path.exists(filename):
    workbook = openpyxl.load_workbook(filename)
    worksheet = workbook.active
else:
    workbook = openpyxl.Workbook()
    worksheet = workbook.active
    headers = ["CID", "UID", "操作者ID", "操作者", "动作", "原因", "时间", "用户名", "组过期时间"]
    worksheet.append(headers)

# 从 Excel 中获取最小的 CID
def get_last_min_cid(worksheet):
    min_cid = float('inf')
    for row in worksheet.iter_rows(min_row=2, values_only=True):
        cid = int(row[0])
        if cid < min_cid:
            min_cid = cid
    return min_cid

current_min_cid = get_last_min_cid(worksheet) if worksheet.max_row > 1 else float('inf')

# 设置基础 URL 和请求参数
base_url = "https://bbs.saraba1st.com/2b/forum.php"
params = {
    "mod": "misc",
    "action": "showdarkroom",
    "ajaxdata": "json",
    "cid": str(current_min_cid) if current_min_cid != float('inf') else None
}

# 最小的 CID 目标
min_target_cid = 5
max_retries = 3  # 最大重试次数

try:
    while current_min_cid > min_target_cid:
        success = False
        for attempt in range(max_retries):
            try:
                response = requests.get(base_url, params=params)
                fixed_json = re.sub(r'(?<=\{|\,)(\s*)(\d+)(\s*):', r'\1"\2"\3:', response.text)
                json_data = json.loads(fixed_json)

                if 'data' in json_data and isinstance(json_data['data'], dict) and json_data['data']:
                    data = json_data['data']
                    fetched_cids = [int(details.get("cid", float('inf'))) for details in data.values()]
                    current_min_cid = min(fetched_cids)
                    params["cid"] = str(current_min_cid)  # 更新 CID 以获取更早的记录
                    success = True

                    for uid, details in data.items():
                        row = [
                            details.get("cid", ""),
                            uid,
                            details.get("operatorid", ""),
                            details.get("operator", ""),
                            details.get("action", ""),
                            details.get("reason", ""),
                            details.get("dateline", ""),
                            details.get("username", ""),
                            details.get("groupexpiry", "")
                        ]
                        worksheet.append(row)

                    print(f"Processed CID down to {current_min_cid}")
                    break
                elif 'data' in json_data and not json_data['data']:  # 检查数据是否为空
                    next_cid = int(json_data['message'].split('|')[1])
                    current_min_cid = next_cid - 1  # 从下一个 CID 继续
                    params["cid"] = str(current_min_cid)
                    print(f"Data missing, continuing from CID {current_min_cid}")
                    success = True
                    break
                else:
                    print("Unexpected data structure received:", json_data)
                    time.sleep(5)
            except json.JSONDecodeError as e:
                print(f"JSON 解析错误: {str(e)}")
                time.sleep(5)
            except Exception as e:
                print(f"请求过程中发生错误: {str(e)}")
                time.sleep(5)
        
        if not success:
            print("重试三次后仍失败，终止程序")
            break

except Exception as e:
    print(f"发生未知错误: {str(e)}")
finally:
    # 保存 Excel 文件
    workbook.save(filename)
    print("数据已成功写入 Excel 文件。")