重构采集脚本并新增按时间导出Excel

- 统一五个站点采集逻辑与启动脚本\n- 新增 dls_fresh 采集流程与日志优化\n- 新增 export_lawyers_excel 按时间条件导出\n- 默认导出近7天并支持扩展字段解析\n- 整理 .gitignore，忽略 data/logs 本地产物

request/requests_client.py

@@ -51,6 +51,7 @@ class RequestsClient:
         self,
         headers: Optional[Mapping[str, str]] = None,
         *,
+        use_proxy: bool = True,
         retry_total: int = 0,
         retry_backoff_factor: float = 0.0,
         retry_status_forcelist: Optional[Iterable[int]] = None,
@@ -58,6 +59,7 @@ class RequestsClient:
         default_timeout: Optional[TimeoutType] = None,
     ) -> None:
         self._base_headers: Dict[str, str] = dict(headers or {})
+        self.use_proxy = bool(use_proxy)
         self.retry_total = int(retry_total)
         self.retry_backoff_factor = float(retry_backoff_factor)
         self.retry_status_forcelist = tuple(retry_status_forcelist or ())
@@ -67,8 +69,13 @@ class RequestsClient:
 
     def _build_session(self) -> requests.Session:
         session = requests.Session()
-        # 统一从 proxy_settings.json 注入代理，并屏蔽系统环境代理干扰
-        apply_proxy(session)
+        if self.use_proxy:
+            # 统一从 proxy_settings.json 注入代理，并屏蔽系统环境代理干扰
+            apply_proxy(session)
+        else:
+            # 强制直连：不读取环境代理，不走配置文件代理
+            session.trust_env = False
+            session.proxies.clear()
         if self.retry_total > 0:
             # 适配器级重试：主要处理连接波动与指定状态码的瞬时失败
             retries = Retry(
@@ -109,6 +116,7 @@ class RequestsClient:
         # 线程场景建议 clone：复用同配置，但使用独立连接池
         clone_client = RequestsClient(
             headers=dict(self.headers),
+            use_proxy=self.use_proxy,
             retry_total=self.retry_total,
             retry_backoff_factor=self.retry_backoff_factor,
             retry_status_forcelist=self.retry_status_forcelist,