# This Python file uses the following encoding: utf-8 # @author runhey """导出脱敏后的配置摘要与最近日志为 zip,用于风控对照分析。不导出账号/token/密码。""" import json import platform import re import sys import zipfile from datetime import datetime from pathlib import Path from module.config.utils import read_file, deep_get from module.server.config_manager import ConfigManager from module.logger import logger PROJECT_ROOT = Path.cwd().resolve() LOG_DIR = PROJECT_ROOT / 'log' OUT_DIR = LOG_DIR / 'diagnostic' _RECENT_INSTANCE_LOGS = 4 _RECENT_SHARED_LOGS = 3 _RECENT_ERROR_LOGS = 10 # 白名单:只导出这些字段,不带出 URS 凭据 _DEVICE_FIELDS = [ 'control_method', 'screenshot_method', 'serial', 'package_name', 'emulatorinfo_type', 'emulatorinfo_name', ] # 值匹配到分隔符为止(涵盖 base64 的 +/= 等), 避免只按 \w 截断导致凭据残留 _V = r'[^\s"\',:;&?{}\[\]]' _SCRUB_PATTERNS = [ (re.compile(r'(账号[::]\s*)(' + _V + r'{3,})'), r'\1***'), (re.compile(r'(account["\'\s::=]+)(' + _V + r'{3,})', re.I), r'\1***'), (re.compile(r'(token["\'\s::=]+)(' + _V + r'{6,})', re.I), r'\1***'), (re.compile(r'(urs[\w]*["\'\s::=]+)(' + _V + r'{6,})', re.I), r'\1***'), (re.compile(r'(password["\'\s::=]+)(' + _V + r'+)', re.I), r'\1***'), (re.compile(r'\b(1[3-9]\d)\d{4}(\d{4})\b'), r'\1****\2'), ] def _scrub(text: str) -> str: for pat, repl in _SCRUB_PATTERNS: text = pat.sub(repl, text) return text def _config_summary(name: str) -> dict: data = read_file(str(PROJECT_ROOT / 'config' / f'{name}.json')) if not isinstance(data, dict): return {'instance': name, 'error': 'config unreadable'} device = {k: deep_get(data, f'script.device.{k}') for k in _DEVICE_FIELDS} optimization = data.get('script', {}).get('optimization', {}) anti_ban = data.get('script', {}).get('anti_ban', {}) enabled = [] for task, cfg in data.items(): if isinstance(cfg, dict) and deep_get(cfg, 'scheduler.enable') is True: enabled.append(task) return { 'instance': name, 'device': device, 'optimization': optimization, 'anti_ban': anti_ban, 'enabled_tasks': sorted(enabled), } def _recent_files(pattern: str, limit: int) -> list: try: candidates = [p for p in LOG_DIR.glob(pattern) if p.is_file()] except OSError: return [] recent = [] for p in candidates: try: recent.append((p.stat().st_mtime, p)) except OSError: continue recent.sort(key=lambda item: item[0], reverse=True) return [p for _, p in recent[:limit]] def build_diagnostic_zip(config_name: str = '') -> Path: """config_name 为空或 'Home' 时导出全部实例。""" # 只接受已存在的实例名, 防止 config_name 带 ../ 造成路径穿越 valid = ConfigManager.all_script_files() if config_name and config_name in valid: instances = [config_name] tag = config_name else: instances = valid tag = 'all' summary = { 'generated_at': datetime.now().strftime('%Y-%m-%d %H:%M:%S'), 'platform': platform.platform(), 'python': sys.version.split()[0], 'instances': [_config_summary(n) for n in instances], } OUT_DIR.mkdir(parents=True, exist_ok=True) stamp = datetime.now().strftime('%Y%m%d_%H%M%S_%f') zip_path = OUT_DIR / f'oas_diag_{tag}_{stamp}.zip' log_files = [] for name in instances: log_files += _recent_files(f'*_{name}.txt', _RECENT_INSTANCE_LOGS) log_files += _recent_files('*_api.txt', _RECENT_SHARED_LOGS) log_files += _recent_files('*_server.txt', _RECENT_SHARED_LOGS) error_logs = [] error_root = LOG_DIR / 'error' if error_root.exists(): try: err = [p for p in error_root.glob('*/log.txt') if p.is_file()] err.sort(key=lambda p: p.stat().st_mtime, reverse=True) error_logs = err[:_RECENT_ERROR_LOGS] except OSError: error_logs = [] with zipfile.ZipFile(zip_path, 'w', zipfile.ZIP_DEFLATED) as zf: zf.writestr('summary.json', json.dumps(summary, ensure_ascii=False, indent=2)) for f in dict.fromkeys(log_files): try: zf.writestr(f'log/{f.name}', _scrub(f.read_text(encoding='utf-8', errors='replace'))) except OSError as e: logger.warning(f'diagnostic: skip {f}: {e}') for f in error_logs: try: zf.writestr(f'error/{f.parent.name}/log.txt', _scrub(f.read_text(encoding='utf-8', errors='replace'))) except OSError as e: logger.warning(f'diagnostic: skip {f}: {e}') logger.info(f'诊断日志已导出: {zip_path}') return zip_path