crawlo 1.2.2__py3-none-any.whl → 1.2.4__py3-none-any.whl

crawlo/downloader/httpx_downloader.py

@@ -1,259 +1,259 @@
-#!/usr/bin/python
-# -*- coding:UTF-8 -*-
-import httpx
-from typing import Optional
-from httpx import AsyncClient, Timeout, Limits
-
-from crawlo.network.response import Response
-from crawlo.downloader import DownloaderBase
-from crawlo.utils.log import get_logger
-
-# 尝试导入 httpx 异常，用于更精确地捕获
-try:
-    # httpx 0.23.0+ 将异常移到了 _exceptions
-    from httpx import ConnectError, TimeoutException, NetworkError, HTTPStatusError
-except ImportError:
-    try:
-        # 旧版本可能在 httpcore 或顶层
-        from httpcore import ConnectError
-        from httpx import TimeoutException, NetworkError, HTTPStatusError
-    except ImportError:
-        ConnectError = httpx.ConnectError
-        TimeoutException = httpx.TimeoutException
-        NetworkError = httpx.NetworkError
-        HTTPStatusError = httpx.HTTPStatusError
-
-# 定义我们认为是网络问题，应该触发降级的异常
-NETWORK_EXCEPTIONS = (ConnectError, TimeoutException, NetworkError)
-
-
-class HttpXDownloader(DownloaderBase):
-    """
-    基于 httpx 的高性能异步下载器
-    - 使用持久化 AsyncClient（推荐做法）
-    - 支持连接池、HTTP/2、透明代理
-    - 智能处理 Request 的 json_body 和 form_data
-    - 支持代理失败后自动降级为直连
-    """
-
-    def __init__(self, crawler):
-        super().__init__(crawler)
-        self._client: Optional[AsyncClient] = None
-        self._client_timeout: Optional[Timeout] = None
-        self._client_limits: Optional[Limits] = None
-        self._client_verify: bool = True
-        self._client_http2: bool = False
-        self.max_download_size: Optional[int] = None
-        # ------------------------
-        self._timeout: Optional[Timeout] = None
-        self._limits: Optional[Limits] = None
-        # --- 获取 logger 实例 ---
-        self.logger = get_logger(self.__class__.__name__, crawler.settings.get("LOG_LEVEL"))
-
-    def open(self):
-        super().open()
-        self.logger.info("Opening HttpXDownloader")
-
-        # 读取配置
-        timeout_total = self.crawler.settings.get_int("DOWNLOAD_TIMEOUT", 30)
-        pool_limit = self.crawler.settings.get_int("CONNECTION_POOL_LIMIT", 100)
-        pool_per_host = self.crawler.settings.get_int("CONNECTION_POOL_LIMIT_PER_HOST", 20)
-        max_download_size = self.crawler.settings.get_int("DOWNLOAD_MAXSIZE", 10 * 1024 * 1024)  # 10MB
-
-        # 保存配置
-        self.max_download_size = max_download_size
-
-        # --- 保存客户端配置以便复用 ---
-        self._client_timeout = Timeout(
-            connect=10.0,  # 建立连接超时
-            read=timeout_total - 10.0 if timeout_total > 10 else timeout_total / 2,  # 读取数据超时
-            write=10.0,  # 发送数据超时
-            pool=1.0  # 从连接池获取连接的超时
-        )
-        self._client_limits = Limits(
-            max_connections=pool_limit,
-            max_keepalive_connections=pool_per_host
-        )
-        self._client_verify = self.crawler.settings.get_bool("VERIFY_SSL", True)
-        self._client_http2 = True  # 启用 HTTP/2 支持
-        # ----------------------------
-
-        # 创建持久化客户端 (不在此处设置全局代理)
-        self._client = AsyncClient(
-            timeout=self._client_timeout,
-            limits=self._client_limits,
-            verify=self._client_verify,
-            http2=self._client_http2,
-            follow_redirects=True,  # 自动跟随重定向
-            # 注意：此处不设置 proxy 或 proxies
-        )
-
-        self.logger.debug("HttpXDownloader initialized.")
-
-    async def download(self, request) -> Optional[Response]:
-        """下载请求并返回响应，支持代理失败后的优雅降级"""
-        if not self._client:
-            raise RuntimeError("HttpXDownloader client is not available.")
-
-        start_time = None
-        if self.crawler.settings.get_bool("DOWNLOAD_STATS", True):
-            import time
-            start_time = time.time()
-
-        # --- 1. 确定要使用的 client 实例 ---
-        effective_client = self._client  # 默认使用共享的主 client
-        temp_client = None  # 用于可能创建的临时 client
-        used_proxy = None  # 记录当前尝试使用的代理
-
-        try:
-            # --- 2. 构造发送参数 (不包含 proxy/proxies) ---
-            kwargs = {
-                "method": request.method,
-                "url": request.url,
-                "headers": request.headers,
-                "cookies": request.cookies,
-                "follow_redirects": request.allow_redirects,
-            }
-
-            # 智能处理 body（关键优化）
-            if hasattr(request, "_json_body") and request._json_body is not None:
-                kwargs["json"] = request._json_body  # 让 httpx 处理序列化
-            elif isinstance(request.body, (dict, list)):
-                kwargs["json"] = request.body
-            else:
-                kwargs["content"] = request.body  # 使用 content 而不是 data
-
-            # --- 3. 处理代理 ---
-            httpx_proxy_config = None  # 用于初始化临时 client 的代理配置
-            if request.proxy:
-                # 根据 request.proxy 的类型准备 httpx 的 proxy 参数
-                if isinstance(request.proxy, str):
-                    # 直接是代理 URL 字符串
-                    httpx_proxy_config = request.proxy
-                elif isinstance(request.proxy, dict):
-                    # 从字典中选择合适的代理 URL
-                    # 优先选择与请求协议匹配的，否则 fallback 到 http
-                    from urllib.parse import urlparse
-                    request_scheme = urlparse(request.url).scheme
-                    if request_scheme == "https" and request.proxy.get("https"):
-                        httpx_proxy_config = request.proxy["https"]
-                    elif request.proxy.get("http"):
-                        httpx_proxy_config = request.proxy["http"]
-                    else:
-                        # 如果没有匹配的，尝试使用任意一个
-                        httpx_proxy_config = next(iter(request.proxy.values()), None)
-                        if httpx_proxy_config:
-                            self.logger.warning(
-                                f"No specific proxy for scheme '{request_scheme}', using '{httpx_proxy_config}'"
-                            )
-
-                # 如果成功解析出代理配置，则创建临时 client
-                if httpx_proxy_config:
-                    try:
-                        # --- 4. 创建临时 client，配置代理 ---
-                        # 使用在 open() 中保存的配置
-                        temp_client = AsyncClient(
-                            timeout=self._client_timeout,
-                            limits=self._client_limits,
-                            verify=self._client_verify,
-                            http2=self._client_http2,
-                            follow_redirects=True,  # 确保继承
-                            proxy=httpx_proxy_config,  # 设置代理
-                        )
-                        effective_client = temp_client
-                        used_proxy = httpx_proxy_config  # 记录使用的代理
-                        self.logger.debug(f"Using temporary client with proxy: {httpx_proxy_config} for {request.url}")
-                    except Exception as e:
-                        self.logger.error(
-                            f"Failed to create temporary client with proxy {httpx_proxy_config} for {request.url}: {e}")
-                        # 出错则回退到使用主 client（无代理）
-                        # 可以选择抛出异常或继续
-                        # raise # 如果希望代理失败导致请求失败，取消注释
-
-            # --- 5. 发送请求 (带降级逻辑) ---
-            try:
-                httpx_response = await effective_client.request(**kwargs)
-            except NETWORK_EXCEPTIONS as proxy_error:
-                # --- 优雅降级逻辑 ---
-                # 如果我们刚刚尝试使用了代理 (temp_client) 并且失败了
-                if temp_client is not None and effective_client is temp_client:
-                    # 记录警告日志
-                    self.logger.warning(
-                        f"代理请求失败 ({used_proxy}), 正在尝试直连: {request.url} | 错误: {repr(proxy_error)}"
-                    )
-                    # 关闭失败的临时客户端
-                    await temp_client.aclose()
-                    temp_client = None  # 防止 finally 再次关闭
-
-                    # 切换到主客户端（直连）
-                    effective_client = self._client
-                    # 再次尝试发送请求
-                    httpx_response = await effective_client.request(**kwargs)
-                else:
-                    # 如果是主客户端（直连）失败，或者不是网络错误，则重新抛出
-                    raise
-
-            # --- 6. 安全检查：防止大响应体 ---
-            content_length = httpx_response.headers.get("Content-Length")
-            if content_length and int(content_length) > self.max_download_size:
-                await httpx_response.aclose()  # 立即关闭连接，释放资源
-                raise OverflowError(f"Response too large: {content_length} > {self.max_download_size}")
-
-            # --- 7. 读取响应体 ---
-            body = await httpx_response.aread()
-
-            # --- 8. 记录下载统计 ---
-            if start_time:
-                download_time = time.time() - start_time
-                self.logger.debug(f"Downloaded {request.url} in {download_time:.3f}s, size: {len(body)} bytes")
-
-            # --- 9. 构造并返回 Response ---
-            return self.structure_response(request=request, response=httpx_response, body=body)
-
-        except httpx.TimeoutException as e:
-            self.logger.error(f"Timeout error for {request.url}: {e}")
-            raise
-        except httpx.NetworkError as e:
-            self.logger.error(f"Network error for {request.url}: {e}")
-            raise
-        except httpx.HTTPStatusError as e:
-            self.logger.warning(f"HTTP {e.response.status_code} for {request.url}: {e}")
-            # 即使是 4xx/5xx，也返回 Response，由上层逻辑（如 spider）处理
-            # 如果需要在此处 raise，可取消注释下一行
-            # raise
-            # 读取响应体以便 structure_response 处理
-            try:
-                error_body = await e.response.aread()
-            except Exception:
-                error_body = b""  # 如果读取错误响应体失败，则为空
-            return self.structure_response(request=request, response=e.response, body=error_body)
-        except Exception as e:
-            self.logger.critical(f"Unexpected error for {request.url}: {e}", exc_info=True)
-            raise
-
-        finally:
-            # --- 10. 清理：关闭临时 client ---
-            # 如果创建了临时 client，则关闭它
-            if temp_client:
-                try:
-                    await temp_client.aclose()
-                    # self.logger.debug("Closed temporary client.")
-                except Exception as e:
-                    self.logger.warning(f"Error closing temporary client: {e}")
-
-    @staticmethod
-    def structure_response(request, response: httpx.Response, body: bytes) -> Response:
-        return Response(
-            url=str(response.url),  # httpx 的 URL 是对象，需转字符串
-            headers=dict(response.headers),
-            status_code=response.status_code,  # 注意：使用 status_code
-            body=body,
-            request=request
-        )
-
-    async def close(self) -> None:
-        """关闭主客户端"""
-        if self._client:
-            self.logger.info("Closing HttpXDownloader client...")
-            await self._client.aclose()
-        self.logger.debug("HttpXDownloader closed.")
+#!/usr/bin/python
+# -*- coding:UTF-8 -*-
+import httpx
+from typing import Optional
+from httpx import AsyncClient, Timeout, Limits
+
+from crawlo.network.response import Response
+from crawlo.downloader import DownloaderBase
+from crawlo.utils.log import get_logger
+
+# 尝试导入 httpx 异常，用于更精确地捕获
+try:
+    # httpx 0.23.0+ 将异常移到了 _exceptions
+    from httpx import ConnectError, TimeoutException, NetworkError, HTTPStatusError
+except ImportError:
+    try:
+        # 旧版本可能在 httpcore 或顶层
+        from httpcore import ConnectError
+        from httpx import TimeoutException, NetworkError, HTTPStatusError
+    except ImportError:
+        ConnectError = httpx.ConnectError
+        TimeoutException = httpx.TimeoutException
+        NetworkError = httpx.NetworkError
+        HTTPStatusError = httpx.HTTPStatusError
+
+# 定义我们认为是网络问题，应该触发降级的异常
+NETWORK_EXCEPTIONS = (ConnectError, TimeoutException, NetworkError)
+
+
+class HttpXDownloader(DownloaderBase):
+    """
+    基于 httpx 的高性能异步下载器
+    - 使用持久化 AsyncClient（推荐做法）
+    - 支持连接池、HTTP/2、透明代理
+    - 智能处理 Request 的 json_body 和 form_data
+    - 支持代理失败后自动降级为直连
+    """
+
+    def __init__(self, crawler):
+        super().__init__(crawler)
+        self._client: Optional[AsyncClient] = None
+        self._client_timeout: Optional[Timeout] = None
+        self._client_limits: Optional[Limits] = None
+        self._client_verify: bool = True
+        self._client_http2: bool = False
+        self.max_download_size: Optional[int] = None
+        # ------------------------
+        self._timeout: Optional[Timeout] = None
+        self._limits: Optional[Limits] = None
+        # --- 获取 logger 实例 ---
+        self.logger = get_logger(self.__class__.__name__, crawler.settings.get("LOG_LEVEL"))
+
+    def open(self):
+        super().open()
+        self.logger.info("Opening HttpXDownloader")
+
+        # 读取配置
+        timeout_total = self.crawler.settings.get_int("DOWNLOAD_TIMEOUT", 30)
+        pool_limit = self.crawler.settings.get_int("CONNECTION_POOL_LIMIT", 100)
+        pool_per_host = self.crawler.settings.get_int("CONNECTION_POOL_LIMIT_PER_HOST", 20)
+        max_download_size = self.crawler.settings.get_int("DOWNLOAD_MAXSIZE", 10 * 1024 * 1024)  # 10MB
+
+        # 保存配置
+        self.max_download_size = max_download_size
+
+        # --- 保存客户端配置以便复用 ---
+        self._client_timeout = Timeout(
+            connect=10.0,  # 建立连接超时
+            read=timeout_total - 10.0 if timeout_total > 10 else timeout_total / 2,  # 读取数据超时
+            write=10.0,  # 发送数据超时
+            pool=1.0  # 从连接池获取连接的超时
+        )
+        self._client_limits = Limits(
+            max_connections=pool_limit,
+            max_keepalive_connections=pool_per_host
+        )
+        self._client_verify = self.crawler.settings.get_bool("VERIFY_SSL", True)
+        self._client_http2 = True  # 启用 HTTP/2 支持
+        # ----------------------------
+
+        # 创建持久化客户端 (不在此处设置全局代理)
+        self._client = AsyncClient(
+            timeout=self._client_timeout,
+            limits=self._client_limits,
+            verify=self._client_verify,
+            http2=self._client_http2,
+            follow_redirects=True,  # 自动跟随重定向
+            # 注意：此处不设置 proxy 或 proxies
+        )
+
+        self.logger.debug("HttpXDownloader initialized.")
+
+    async def download(self, request) -> Optional[Response]:
+        """下载请求并返回响应，支持代理失败后的优雅降级"""
+        if not self._client:
+            raise RuntimeError("HttpXDownloader client is not available.")
+
+        start_time = None
+        if self.crawler.settings.get_bool("DOWNLOAD_STATS", True):
+            import time
+            start_time = time.time()
+
+        # --- 1. 确定要使用的 client 实例 ---
+        effective_client = self._client  # 默认使用共享的主 client
+        temp_client = None  # 用于可能创建的临时 client
+        used_proxy = None  # 记录当前尝试使用的代理
+
+        try:
+            # --- 2. 构造发送参数 (不包含 proxy/proxies) ---
+            kwargs = {
+                "method": request.method,
+                "url": request.url,
+                "headers": request.headers,
+                "cookies": request.cookies,
+                "follow_redirects": request.allow_redirects,
+            }
+
+            # 智能处理 body（关键优化）
+            if hasattr(request, "_json_body") and request._json_body is not None:
+                kwargs["json"] = request._json_body  # 让 httpx 处理序列化
+            elif isinstance(request.body, (dict, list)):
+                kwargs["json"] = request.body
+            else:
+                kwargs["content"] = request.body  # 使用 content 而不是 data
+
+            # --- 3. 处理代理 ---
+            httpx_proxy_config = None  # 用于初始化临时 client 的代理配置
+            if request.proxy:
+                # 根据 request.proxy 的类型准备 httpx 的 proxy 参数
+                if isinstance(request.proxy, str):
+                    # 直接是代理 URL 字符串
+                    httpx_proxy_config = request.proxy
+                elif isinstance(request.proxy, dict):
+                    # 从字典中选择合适的代理 URL
+                    # 优先选择与请求协议匹配的，否则 fallback 到 http
+                    from urllib.parse import urlparse
+                    request_scheme = urlparse(request.url).scheme
+                    if request_scheme == "https" and request.proxy.get("https"):
+                        httpx_proxy_config = request.proxy["https"]
+                    elif request.proxy.get("http"):
+                        httpx_proxy_config = request.proxy["http"]
+                    else:
+                        # 如果没有匹配的，尝试使用任意一个
+                        httpx_proxy_config = next(iter(request.proxy.values()), None)
+                        if httpx_proxy_config:
+                            self.logger.warning(
+                                f"No specific proxy for scheme '{request_scheme}', using '{httpx_proxy_config}'"
+                            )
+
+                # 如果成功解析出代理配置，则创建临时 client
+                if httpx_proxy_config:
+                    try:
+                        # --- 4. 创建临时 client，配置代理 ---
+                        # 使用在 open() 中保存的配置
+                        temp_client = AsyncClient(
+                            timeout=self._client_timeout,
+                            limits=self._client_limits,
+                            verify=self._client_verify,
+                            http2=self._client_http2,
+                            follow_redirects=True,  # 确保继承
+                            proxy=httpx_proxy_config,  # 设置代理
+                        )
+                        effective_client = temp_client
+                        used_proxy = httpx_proxy_config  # 记录使用的代理
+                        self.logger.debug(f"Using temporary client with proxy: {httpx_proxy_config} for {request.url}")
+                    except Exception as e:
+                        self.logger.error(
+                            f"Failed to create temporary client with proxy {httpx_proxy_config} for {request.url}: {e}")
+                        # 出错则回退到使用主 client（无代理）
+                        # 可以选择抛出异常或继续
+                        # raise # 如果希望代理失败导致请求失败，取消注释
+
+            # --- 5. 发送请求 (带降级逻辑) ---
+            try:
+                httpx_response = await effective_client.request(**kwargs)
+            except NETWORK_EXCEPTIONS as proxy_error:
+                # --- 优雅降级逻辑 ---
+                # 如果我们刚刚尝试使用了代理 (temp_client) 并且失败了
+                if temp_client is not None and effective_client is temp_client:
+                    # 记录警告日志
+                    self.logger.warning(
+                        f"代理请求失败 ({used_proxy}), 正在尝试直连: {request.url} | 错误: {repr(proxy_error)}"
+                    )
+                    # 关闭失败的临时客户端
+                    await temp_client.aclose()
+                    temp_client = None  # 防止 finally 再次关闭
+
+                    # 切换到主客户端（直连）
+                    effective_client = self._client
+                    # 再次尝试发送请求
+                    httpx_response = await effective_client.request(**kwargs)
+                else:
+                    # 如果是主客户端（直连）失败，或者不是网络错误，则重新抛出
+                    raise
+
+            # --- 6. 安全检查：防止大响应体 ---
+            content_length = httpx_response.headers.get("Content-Length")
+            if content_length and int(content_length) > self.max_download_size:
+                await httpx_response.aclose()  # 立即关闭连接，释放资源
+                raise OverflowError(f"Response too large: {content_length} > {self.max_download_size}")
+
+            # --- 7. 读取响应体 ---
+            body = await httpx_response.aread()
+
+            # --- 8. 记录下载统计 ---
+            if start_time:
+                download_time = time.time() - start_time
+                self.logger.debug(f"Downloaded {request.url} in {download_time:.3f}s, size: {len(body)} bytes")
+
+            # --- 9. 构造并返回 Response ---
+            return self.structure_response(request=request, response=httpx_response, body=body)
+
+        except httpx.TimeoutException as e:
+            self.logger.error(f"Timeout error for {request.url}: {e}")
+            raise
+        except httpx.NetworkError as e:
+            self.logger.error(f"Network error for {request.url}: {e}")
+            raise
+        except httpx.HTTPStatusError as e:
+            self.logger.warning(f"HTTP {e.response.status_code} for {request.url}: {e}")
+            # 即使是 4xx/5xx，也返回 Response，由上层逻辑（如 spider）处理
+            # 如果需要在此处 raise，可取消注释下一行
+            # raise
+            # 读取响应体以便 structure_response 处理
+            try:
+                error_body = await e.response.aread()
+            except Exception:
+                error_body = b""  # 如果读取错误响应体失败，则为空
+            return self.structure_response(request=request, response=e.response, body=error_body)
+        except Exception as e:
+            self.logger.critical(f"Unexpected error for {request.url}: {e}", exc_info=True)
+            raise
+
+        finally:
+            # --- 10. 清理：关闭临时 client ---
+            # 如果创建了临时 client，则关闭它
+            if temp_client:
+                try:
+                    await temp_client.aclose()
+                    # self.logger.debug("Closed temporary client.")
+                except Exception as e:
+                    self.logger.warning(f"Error closing temporary client: {e}")
+
+    @staticmethod
+    def structure_response(request, response: httpx.Response, body: bytes) -> Response:
+        return Response(
+            url=str(response.url),  # httpx 的 URL 是对象，需转字符串
+            headers=dict(response.headers),
+            status_code=response.status_code,  # 注意：使用 status_code
+            body=body,
+            request=request
+        )
+
+    async def close(self) -> None:
+        """关闭主客户端"""
+        if self._client:
+            self.logger.info("Closing HttpXDownloader client...")
+            await self._client.aclose()
+        self.logger.debug("HttpXDownloader closed.")