MCP Development Framework

MIT License
OverviewInspectSchema Related Servers Reviews Score
{
    "sourceFile": "mcp_tool/tools/image_recognition.py",
    "activeCommit": 0,
    "commits": [
        {
            "activePatchIndex": 4,
            "patches": [
                {
                    "date": 1741521023660,
                    "content": "Index: \n===================================================================\n--- \n+++ \n"
                },
                {
                    "date": 1741521301820,
                    "content": "Index: \n===================================================================\n--- \n+++ \n@@ -13,8 +13,14 @@\n # 配置日志\n logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')\n logger = logging.getLogger(__name__)\n \n+# 路径映射配置\n+# 本地路径到API服务路径的映射\n+PATH_MAPPING = {\n+    '/host_images': os.environ.get('API_IMAGE_PATH', '/host_images')\n+}\n+\n class ImageRecognitionClient:\n     \"\"\"\n     图像识别客户端，用于调用外部大模型API进行图像识别\n     \"\"\"\n@@ -59,8 +65,11 @@\n         if not self.api_key:\n             logger.error(\"API密钥未设置，无法进行图像识别\")\n             return None\n         \n+        # 映射路径，处理不同容器间的路径差异\n+        mapped_path = self._map_path(image_path)\n+        \n         if not os.path.exists(image_path):\n             logger.error(f\"图像文件不存在: {image_path}\")\n             return None\n         \n@@ -80,9 +89,9 @@\n                         },\n                         {\n                             \"type\": \"image_url\",\n                             \"image_url\": {\n-                                \"url\": self._get_image_url(image_path)\n+                                \"url\": self._get_image_url(mapped_path)\n                             }\n                         }\n                     ]\n                 }\n@@ -128,8 +137,32 @@\n         except Exception as e:\n             logger.error(f\"图像识别过程中发生错误: {str(e)}\")\n             return None\n     \n+    def _map_path(self, image_path: str) -> str:\n+        \"\"\"\n+        映射路径，处理不同容器间的路径差异\n+        \n+        Args:\n+            image_path: 原始图像路径\n+            \n+        Returns:\n+            映射后的路径\n+        \"\"\"\n+        # 如果是URL，不需要映射\n+        if image_path.startswith(('http://', 'https://')):\n+            return image_path\n+            \n+        # 检查是否需要路径映射\n+        for local_path, api_path in PATH_MAPPING.items():\n+            if image_path.startswith(local_path):\n+                mapped_path = image_path.replace(local_path, api_path)\n+                logger.info(f\"路径映射: {image_path} -> {mapped_path}\")\n+                return mapped_path\n+        \n+        # 如果没有匹配的映射，返回原路径\n+        return image_path\n+    \n     def _get_image_url(self, image_path: str) -> str:\n         \"\"\"\n         获取图像URL，如果是本地文件则转换为base64编码\n         \n"
                },
                {
                    "date": 1741522010479,
                    "content": "Index: \n===================================================================\n--- \n+++ \n@@ -6,8 +6,10 @@\n import http.client\n import json\n import base64\n import logging\n+import ssl\n+import time\n from typing import Optional, Dict, Any, Union\n from urllib.parse import urlparse\n \n # 配置日志\n@@ -19,36 +21,53 @@\n PATH_MAPPING = {\n     '/host_images': os.environ.get('API_IMAGE_PATH', '/host_images')\n }\n \n+# 超时设置\n+DEFAULT_TIMEOUT = int(os.environ.get('LLM_API_TIMEOUT', '60'))  # 默认60秒超时\n+MAX_RETRIES = int(os.environ.get('LLM_API_MAX_RETRIES', '2'))  # 默认最多重试2次\n+\n class ImageRecognitionClient:\n     \"\"\"\n     图像识别客户端，用于调用外部大模型API进行图像识别\n     \"\"\"\n     \n     def __init__(self, \n                  base_url: str = None, \n                  api_key: str = None, \n-                 model: str = None):\n+                 model: str = None,\n+                 timeout: int = None,\n+                 max_retries: int = None):\n         \"\"\"\n         初始化图像识别客户端\n         \n         Args:\n             base_url: API基础URL，如果为None则从环境变量获取\n             api_key: API密钥，如果为None则从环境变量获取\n             model: 使用的模型名称，如果为None则从环境变量获取\n+            timeout: 请求超时时间（秒），如果为None则从环境变量获取\n+            max_retries: 最大重试次数，如果为None则从环境变量获取\n         \"\"\"\n         # 从环境变量或参数获取配置\n         self.base_url = base_url or os.environ.get('LLM_API_BASE_URL', 'api.openai.com')\n         self.api_key = api_key or os.environ.get('LLM_API_KEY', '')\n         self.model = model or os.environ.get('LLM_MODEL', 'gpt-4o')\n         self.max_tokens = int(os.environ.get('LLM_MAX_TOKENS', '400'))\n+        self.timeout = timeout or DEFAULT_TIMEOUT\n+        self.max_retries = max_retries or MAX_RETRIES\n         \n         # 解析base_url，分离主机名和路径\n-        parsed_url = urlparse(f\"https://{self.base_url}\" if not self.base_url.startswith(('http://', 'https://')) else self.base_url)\n+        parsed_url = urlparse(self.base_url)\n+        # 如果没有协议，默认使用https\n+        if not parsed_url.scheme:\n+            parsed_url = urlparse(f\"https://{self.base_url}\")\n+            \n         self.host = parsed_url.netloc\n         self.base_path = parsed_url.path.rstrip('/')\n+        self.use_https = parsed_url.scheme == 'https'\n         \n+        logger.info(f\"API配置: 主机={self.host}, 路径={self.base_path}, HTTPS={self.use_https}, 超时={self.timeout}秒, 最大重试次数={self.max_retries}\")\n+        \n         # 验证配置\n         if not self.api_key:\n             logger.warning(\"API密钥未设置，图像识别功能将不可用\")\n     \n@@ -72,72 +91,114 @@\n         if not os.path.exists(image_path):\n             logger.error(f\"图像文件不存在: {image_path}\")\n             return None\n         \n-        try:\n-            # 检查是否是图像文件\n-            if not image_path.lower().endswith(('.png', '.jpg', '.jpeg', '.gif', '.bmp', '.webp')):\n-                logger.warning(f\"文件可能不是图像: {image_path}\")\n-            \n-            # 准备请求数据\n-            messages = [\n-                {\n-                    \"role\": \"user\",\n-                    \"content\": [\n-                        {\n-                            \"type\": \"text\",\n-                            \"text\": \"请详细描述这张图片中的内容，包括图片中的文字、图形、颜色等元素。\"\n-                        },\n-                        {\n-                            \"type\": \"image_url\",\n-                            \"image_url\": {\n-                                \"url\": self._get_image_url(mapped_path)\n+        # 重试机制\n+        retries = 0\n+        last_error = None\n+        \n+        while retries <= self.max_retries:\n+            try:\n+                # 检查是否是图像文件\n+                if not image_path.lower().endswith(('.png', '.jpg', '.jpeg', '.gif', '.bmp', '.webp')):\n+                    logger.warning(f\"文件可能不是图像: {image_path}\")\n+                \n+                # 准备请求数据\n+                messages = [\n+                    {\n+                        \"role\": \"user\",\n+                        \"content\": [\n+                            {\n+                                \"type\": \"text\",\n+                                \"text\": \"请详细描述这张图片中的内容，包括图片中的文字、图形、颜色等元素。\"\n+                            },\n+                            {\n+                                \"type\": \"image_url\",\n+                                \"image_url\": {\n+                                    \"url\": self._get_image_url(mapped_path)\n+                                }\n                             }\n-                        }\n-                    ]\n+                        ]\n+                    }\n+                ]\n+                \n+                payload = json.dumps({\n+                    \"model\": self.model,\n+                    \"stream\": False,\n+                    \"messages\": messages,\n+                    \"max_tokens\": self.max_tokens\n+                })\n+                \n+                headers = {\n+                    'Accept': 'application/json',\n+                    'Authorization': f'Bearer {self.api_key}',\n+                    'Content-Type': 'application/json'\n                 }\n-            ]\n-            \n-            payload = json.dumps({\n-                \"model\": self.model,\n-                \"stream\": False,\n-                \"messages\": messages,\n-                \"max_tokens\": self.max_tokens\n-            })\n-            \n-            headers = {\n-                'Accept': 'application/json',\n-                'Authorization': f'Bearer {self.api_key}',\n-                'Content-Type': 'application/json'\n-            }\n-            \n-            # 建立连接并发送请求\n-            conn = http.client.HTTPSConnection(self.host)\n-            conn.request(\"POST\", f\"{self.base_path}/v1/chat/completions\", payload, headers)\n-            \n-            # 获取响应\n-            response = conn.getresponse()\n-            data = response.read().decode(\"utf-8\")\n-            conn.close()\n-            \n-            # 解析响应\n-            response_data = json.loads(data)\n-            \n-            if 'error' in response_data:\n-                logger.error(f\"API错误: {response_data['error']['message']}\")\n+                \n+                # 建立连接并发送请求\n+                if self.use_https:\n+                    # 创建SSL上下文\n+                    context = ssl.create_default_context()\n+                    conn = http.client.HTTPSConnection(self.host, context=context, timeout=self.timeout)\n+                else:\n+                    conn = http.client.HTTPConnection(self.host, timeout=self.timeout)\n+                    \n+                logger.info(f\"发送请求到: {self.host}{self.base_path}/chat/completions (尝试 {retries+1}/{self.max_retries+1})\")\n+                conn.request(\"POST\", f\"{self.base_path}/chat/completions\", payload, headers)\n+                \n+                # 获取响应\n+                response = conn.getresponse()\n+                data = response.read().decode(\"utf-8\")\n+                conn.close()\n+                \n+                logger.info(f\"API响应状态码: {response.status}\")\n+                \n+                # 解析响应\n+                response_data = json.loads(data)\n+                \n+                if 'error' in response_data:\n+                    error_msg = response_data['error']['message'] if 'message' in response_data['error'] else response_data['error']\n+                    logger.error(f\"API错误: {error_msg}\")\n+                    last_error = error_msg\n+                    retries += 1\n+                    if retries <= self.max_retries:\n+                        # 指数退避重试\n+                        wait_time = 2 ** retries\n+                        logger.info(f\"等待 {wait_time} 秒后重试...\")\n+                        time.sleep(wait_time)\n+                        continue\n+                    return None\n+                \n+                # 提取识别结果\n+                if 'choices' in response_data and len(response_data['choices']) > 0:\n+                    result = response_data['choices'][0]['message']['content']\n+                    return result\n+                else:\n+                    logger.error(f\"API响应中未找到识别结果: {response_data}\")\n+                    last_error = \"API响应中未找到识别结果\"\n+                    retries += 1\n+                    if retries <= self.max_retries:\n+                        wait_time = 2 ** retries\n+                        logger.info(f\"等待 {wait_time} 秒后重试...\")\n+                        time.sleep(wait_time)\n+                        continue\n+                    return None\n+                \n+            except Exception as e:\n+                logger.error(f\"图像识别过程中发生错误: {str(e)}\")\n+                import traceback\n+                logger.error(traceback.format_exc())\n+                last_error = str(e)\n+                retries += 1\n+                if retries <= self.max_retries:\n+                    wait_time = 2 ** retries\n+                    logger.info(f\"等待 {wait_time} 秒后重试...\")\n+                    time.sleep(wait_time)\n+                    continue\n                 return None\n-            \n-            # 提取识别结果\n-            if 'choices' in response_data and len(response_data['choices']) > 0:\n-                result = response_data['choices'][0]['message']['content']\n-                return result\n-            else:\n-                logger.error(\"API响应中未找到识别结果\")\n-                return None\n-            \n-        except Exception as e:\n-            logger.error(f\"图像识别过程中发生错误: {str(e)}\")\n-            return None\n+        \n+        logger.error(f\"图像识别失败，已达到最大重试次数 ({self.max_retries})。最后错误: {last_error}\")\n+        return None\n     \n     def _map_path(self, image_path: str) -> str:\n         \"\"\"\n         映射路径，处理不同容器间的路径差异\n"
                },
                {
                    "date": 1741522919675,
                    "content": "Index: \n===================================================================\n--- \n+++ \n@@ -25,8 +25,14 @@\n # 超时设置\n DEFAULT_TIMEOUT = int(os.environ.get('LLM_API_TIMEOUT', '60'))  # 默认60秒超时\n MAX_RETRIES = int(os.environ.get('LLM_API_MAX_RETRIES', '2'))  # 默认最多重试2次\n \n+# 打印当前环境变量值，用于调试\n+logger.info(f\"环境变量中的超时设置: LLM_API_TIMEOUT={os.environ.get('LLM_API_TIMEOUT', '未设置')}\")\n+logger.info(f\"环境变量中的重试设置: LLM_API_MAX_RETRIES={os.environ.get('LLM_API_MAX_RETRIES', '未设置')}\")\n+logger.info(f\"实际使用的超时设置: DEFAULT_TIMEOUT={DEFAULT_TIMEOUT}\")\n+logger.info(f\"实际使用的重试设置: MAX_RETRIES={MAX_RETRIES}\")\n+\n class ImageRecognitionClient:\n     \"\"\"\n     图像识别客户端，用于调用外部大模型API进行图像识别\n     \"\"\"\n@@ -51,11 +57,39 @@\n         self.base_url = base_url or os.environ.get('LLM_API_BASE_URL', 'api.openai.com')\n         self.api_key = api_key or os.environ.get('LLM_API_KEY', '')\n         self.model = model or os.environ.get('LLM_MODEL', 'gpt-4o')\n         self.max_tokens = int(os.environ.get('LLM_MAX_TOKENS', '400'))\n-        self.timeout = timeout or DEFAULT_TIMEOUT\n-        self.max_retries = max_retries or MAX_RETRIES\n         \n+        # 确保正确读取超时和重试设置\n+        timeout_str = os.environ.get('LLM_API_TIMEOUT')\n+        max_retries_str = os.environ.get('LLM_API_MAX_RETRIES')\n+        \n+        # 如果参数提供了值，优先使用参数值\n+        if timeout is not None:\n+            self.timeout = timeout\n+        # 否则尝试从环境变量获取\n+        elif timeout_str is not None:\n+            try:\n+                self.timeout = int(timeout_str)\n+            except ValueError:\n+                logger.warning(f\"环境变量LLM_API_TIMEOUT值无效: {timeout_str}，使用默认值: {DEFAULT_TIMEOUT}\")\n+                self.timeout = DEFAULT_TIMEOUT\n+        # 如果都没有，使用默认值\n+        else:\n+            self.timeout = DEFAULT_TIMEOUT\n+            \n+        # 同样处理重试次数\n+        if max_retries is not None:\n+            self.max_retries = max_retries\n+        elif max_retries_str is not None:\n+            try:\n+                self.max_retries = int(max_retries_str)\n+            except ValueError:\n+                logger.warning(f\"环境变量LLM_API_MAX_RETRIES值无效: {max_retries_str}，使用默认值: {MAX_RETRIES}\")\n+                self.max_retries = MAX_RETRIES\n+        else:\n+            self.max_retries = MAX_RETRIES\n+        \n         # 解析base_url，分离主机名和路径\n         parsed_url = urlparse(self.base_url)\n         # 如果没有协议，默认使用https\n         if not parsed_url.scheme:\n"
                },
                {
                    "date": 1741523194318,
                    "content": "Index: \n===================================================================\n--- \n+++ \n@@ -1,321 +1,1 @@\n-\"\"\"\n-图像识别内部接口模块，用于调用外部大模型进行图像识别\n-\"\"\"\n-\n-import os\n-import http.client\n-import json\n-import base64\n-import logging\n-import ssl\n-import time\n-from typing import Optional, Dict, Any, Union\n-from urllib.parse import urlparse\n-\n-# 配置日志\n-logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')\n-logger = logging.getLogger(__name__)\n-\n-# 路径映射配置\n-# 本地路径到API服务路径的映射\n-PATH_MAPPING = {\n-    '/host_images': os.environ.get('API_IMAGE_PATH', '/host_images')\n-}\n-\n-# 超时设置\n-DEFAULT_TIMEOUT = int(os.environ.get('LLM_API_TIMEOUT', '60'))  # 默认60秒超时\n-MAX_RETRIES = int(os.environ.get('LLM_API_MAX_RETRIES', '2'))  # 默认最多重试2次\n-\n-# 打印当前环境变量值，用于调试\n-logger.info(f\"环境变量中的超时设置: LLM_API_TIMEOUT={os.environ.get('LLM_API_TIMEOUT', '未设置')}\")\n-logger.info(f\"环境变量中的重试设置: LLM_API_MAX_RETRIES={os.environ.get('LLM_API_MAX_RETRIES', '未设置')}\")\n-logger.info(f\"实际使用的超时设置: DEFAULT_TIMEOUT={DEFAULT_TIMEOUT}\")\n-logger.info(f\"实际使用的重试设置: MAX_RETRIES={MAX_RETRIES}\")\n-\n-class ImageRecognitionClient:\n-    \"\"\"\n-    图像识别客户端，用于调用外部大模型API进行图像识别\n-    \"\"\"\n-    \n-    def __init__(self, \n-                 base_url: str = None, \n-                 api_key: str = None, \n-                 model: str = None,\n-                 timeout: int = None,\n-                 max_retries: int = None):\n-        \"\"\"\n-        初始化图像识别客户端\n-        \n-        Args:\n-            base_url: API基础URL，如果为None则从环境变量获取\n-            api_key: API密钥，如果为None则从环境变量获取\n-            model: 使用的模型名称，如果为None则从环境变量获取\n-            timeout: 请求超时时间（秒），如果为None则从环境变量获取\n-            max_retries: 最大重试次数，如果为None则从环境变量获取\n-        \"\"\"\n-        # 从环境变量或参数获取配置\n-        self.base_url = base_url or os.environ.get('LLM_API_BASE_URL', 'api.openai.com')\n-        self.api_key = api_key or os.environ.get('LLM_API_KEY', '')\n-        self.model = model or os.environ.get('LLM_MODEL', 'gpt-4o')\n-        self.max_tokens = int(os.environ.get('LLM_MAX_TOKENS', '400'))\n-        \n-        # 确保正确读取超时和重试设置\n-        timeout_str = os.environ.get('LLM_API_TIMEOUT')\n-        max_retries_str = os.environ.get('LLM_API_MAX_RETRIES')\n-        \n-        # 如果参数提供了值，优先使用参数值\n-        if timeout is not None:\n-            self.timeout = timeout\n-        # 否则尝试从环境变量获取\n-        elif timeout_str is not None:\n-            try:\n-                self.timeout = int(timeout_str)\n-            except ValueError:\n-                logger.warning(f\"环境变量LLM_API_TIMEOUT值无效: {timeout_str}，使用默认值: {DEFAULT_TIMEOUT}\")\n-                self.timeout = DEFAULT_TIMEOUT\n-        # 如果都没有，使用默认值\n-        else:\n-            self.timeout = DEFAULT_TIMEOUT\n-            \n-        # 同样处理重试次数\n-        if max_retries is not None:\n-            self.max_retries = max_retries\n-        elif max_retries_str is not None:\n-            try:\n-                self.max_retries = int(max_retries_str)\n-            except ValueError:\n-                logger.warning(f\"环境变量LLM_API_MAX_RETRIES值无效: {max_retries_str}，使用默认值: {MAX_RETRIES}\")\n-                self.max_retries = MAX_RETRIES\n-        else:\n-            self.max_retries = MAX_RETRIES\n-        \n-        # 解析base_url，分离主机名和路径\n-        parsed_url = urlparse(self.base_url)\n-        # 如果没有协议，默认使用https\n-        if not parsed_url.scheme:\n-            parsed_url = urlparse(f\"https://{self.base_url}\")\n-            \n-        self.host = parsed_url.netloc\n-        self.base_path = parsed_url.path.rstrip('/')\n-        self.use_https = parsed_url.scheme == 'https'\n-        \n-        logger.info(f\"API配置: 主机={self.host}, 路径={self.base_path}, HTTPS={self.use_https}, 超时={self.timeout}秒, 最大重试次数={self.max_retries}\")\n-        \n-        # 验证配置\n-        if not self.api_key:\n-            logger.warning(\"API密钥未设置，图像识别功能将不可用\")\n-    \n-    def recognize_image(self, image_path: str) -> Optional[str]:\n-        \"\"\"\n-        识别图像内容\n-        \n-        Args:\n-            image_path: 图像文件路径\n-            \n-        Returns:\n-            识别结果文本，如果识别失败则返回None\n-        \"\"\"\n-        if not self.api_key:\n-            logger.error(\"API密钥未设置，无法进行图像识别\")\n-            return None\n-        \n-        # 映射路径，处理不同容器间的路径差异\n-        mapped_path = self._map_path(image_path)\n-        \n-        if not os.path.exists(image_path):\n-            logger.error(f\"图像文件不存在: {image_path}\")\n-            return None\n-        \n-        # 重试机制\n-        retries = 0\n-        last_error = None\n-        \n-        while retries <= self.max_retries:\n-            try:\n-                # 检查是否是图像文件\n-                if not image_path.lower().endswith(('.png', '.jpg', '.jpeg', '.gif', '.bmp', '.webp')):\n-                    logger.warning(f\"文件可能不是图像: {image_path}\")\n-                \n-                # 准备请求数据\n-                messages = [\n-                    {\n-                        \"role\": \"user\",\n-                        \"content\": [\n-                            {\n-                                \"type\": \"text\",\n-                                \"text\": \"请详细描述这张图片中的内容，包括图片中的文字、图形、颜色等元素。\"\n-                            },\n-                            {\n-                                \"type\": \"image_url\",\n-                                \"image_url\": {\n-                                    \"url\": self._get_image_url(mapped_path)\n-                                }\n-                            }\n-                        ]\n-                    }\n-                ]\n-                \n-                payload = json.dumps({\n-                    \"model\": self.model,\n-                    \"stream\": False,\n-                    \"messages\": messages,\n-                    \"max_tokens\": self.max_tokens\n-                })\n-                \n-                headers = {\n-                    'Accept': 'application/json',\n-                    'Authorization': f'Bearer {self.api_key}',\n-                    'Content-Type': 'application/json'\n-                }\n-                \n-                # 建立连接并发送请求\n-                if self.use_https:\n-                    # 创建SSL上下文\n-                    context = ssl.create_default_context()\n-                    conn = http.client.HTTPSConnection(self.host, context=context, timeout=self.timeout)\n-                else:\n-                    conn = http.client.HTTPConnection(self.host, timeout=self.timeout)\n-                    \n-                logger.info(f\"发送请求到: {self.host}{self.base_path}/chat/completions (尝试 {retries+1}/{self.max_retries+1})\")\n-                conn.request(\"POST\", f\"{self.base_path}/chat/completions\", payload, headers)\n-                \n-                # 获取响应\n-                response = conn.getresponse()\n-                data = response.read().decode(\"utf-8\")\n-                conn.close()\n-                \n-                logger.info(f\"API响应状态码: {response.status}\")\n-                \n-                # 解析响应\n-                response_data = json.loads(data)\n-                \n-                if 'error' in response_data:\n-                    error_msg = response_data['error']['message'] if 'message' in response_data['error'] else response_data['error']\n-                    logger.error(f\"API错误: {error_msg}\")\n-                    last_error = error_msg\n-                    retries += 1\n-                    if retries <= self.max_retries:\n-                        # 指数退避重试\n-                        wait_time = 2 ** retries\n-                        logger.info(f\"等待 {wait_time} 秒后重试...\")\n-                        time.sleep(wait_time)\n-                        continue\n-                    return None\n-                \n-                # 提取识别结果\n-                if 'choices' in response_data and len(response_data['choices']) > 0:\n-                    result = response_data['choices'][0]['message']['content']\n-                    return result\n-                else:\n-                    logger.error(f\"API响应中未找到识别结果: {response_data}\")\n-                    last_error = \"API响应中未找到识别结果\"\n-                    retries += 1\n-                    if retries <= self.max_retries:\n-                        wait_time = 2 ** retries\n-                        logger.info(f\"等待 {wait_time} 秒后重试...\")\n-                        time.sleep(wait_time)\n-                        continue\n-                    return None\n-                \n-            except Exception as e:\n-                logger.error(f\"图像识别过程中发生错误: {str(e)}\")\n-                import traceback\n-                logger.error(traceback.format_exc())\n-                last_error = str(e)\n-                retries += 1\n-                if retries <= self.max_retries:\n-                    wait_time = 2 ** retries\n-                    logger.info(f\"等待 {wait_time} 秒后重试...\")\n-                    time.sleep(wait_time)\n-                    continue\n-                return None\n-        \n-        logger.error(f\"图像识别失败，已达到最大重试次数 ({self.max_retries})。最后错误: {last_error}\")\n-        return None\n-    \n-    def _map_path(self, image_path: str) -> str:\n-        \"\"\"\n-        映射路径，处理不同容器间的路径差异\n-        \n-        Args:\n-            image_path: 原始图像路径\n-            \n-        Returns:\n-            映射后的路径\n-        \"\"\"\n-        # 如果是URL，不需要映射\n-        if image_path.startswith(('http://', 'https://')):\n-            return image_path\n-            \n-        # 检查是否需要路径映射\n-        for local_path, api_path in PATH_MAPPING.items():\n-            if image_path.startswith(local_path):\n-                mapped_path = image_path.replace(local_path, api_path)\n-                logger.info(f\"路径映射: {image_path} -> {mapped_path}\")\n-                return mapped_path\n-        \n-        # 如果没有匹配的映射，返回原路径\n-        return image_path\n-    \n-    def _get_image_url(self, image_path: str) -> str:\n-        \"\"\"\n-        获取图像URL，如果是本地文件则转换为base64编码\n-        \n-        Args:\n-            image_path: 图像文件路径\n-            \n-        Returns:\n-            图像URL或base64编码的数据URL\n-        \"\"\"\n-        # 如果是URL则直接返回\n-        if image_path.startswith(('http://', 'https://')):\n-            return image_path\n-        \n-        # 否则将本地文件转换为base64编码的数据URL\n-        try:\n-            with open(image_path, \"rb\") as image_file:\n-                encoded_string = base64.b64encode(image_file.read()).decode('utf-8')\n-            \n-            # 确定MIME类型\n-            mime_type = self._get_mime_type(image_path)\n-            return f\"data:{mime_type};base64,{encoded_string}\"\n-        except Exception as e:\n-            logger.error(f\"图像文件读取或编码失败: {str(e)}\")\n-            raise\n-    \n-    def _get_mime_type(self, file_path: str) -> str:\n-        \"\"\"\n-        根据文件扩展名获取MIME类型\n-        \n-        Args:\n-            file_path: 文件路径\n-            \n-        Returns:\n-            MIME类型字符串\n-        \"\"\"\n-        extension = os.path.splitext(file_path)[1].lower()\n-        mime_types = {\n-            '.jpg': 'image/jpeg',\n-            '.jpeg': 'image/jpeg',\n-            '.png': 'image/png',\n-            '.gif': 'image/gif',\n-            '.bmp': 'image/bmp',\n-            '.webp': 'image/webp'\n-        }\n-        return mime_types.get(extension, 'application/octet-stream')\n-\n-\n-# 创建默认客户端实例，方便直接导入使用\n-default_client = ImageRecognitionClient()\n-\n-def recognize_image(image_path: str) -> Optional[str]:\n-    \"\"\"\n-    使用默认客户端识别图像内容\n-    \n-    Args:\n-        image_path: 图像文件路径\n-        \n-    Returns:\n-        识别结果文本，如果识别失败则返回None\n-    \"\"\"\n-    return default_client.recognize_image(image_path) \n\\ No newline at end of file\n+ \n\\ No newline at end of file\n"
                }
            ],
            "date": 1741521023660,
            "name": "Commit-0",
            "content": "\"\"\"\n图像识别内部接口模块，用于调用外部大模型进行图像识别\n\"\"\"\n\nimport os\nimport http.client\nimport json\nimport base64\nimport logging\nfrom typing import Optional, Dict, Any, Union\nfrom urllib.parse import urlparse\n\n# 配置日志\nlogging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')\nlogger = logging.getLogger(__name__)\n\nclass ImageRecognitionClient:\n    \"\"\"\n    图像识别客户端，用于调用外部大模型API进行图像识别\n    \"\"\"\n    \n    def __init__(self, \n                 base_url: str = None, \n                 api_key: str = None, \n                 model: str = None):\n        \"\"\"\n        初始化图像识别客户端\n        \n        Args:\n            base_url: API基础URL，如果为None则从环境变量获取\n            api_key: API密钥，如果为None则从环境变量获取\n            model: 使用的模型名称，如果为None则从环境变量获取\n        \"\"\"\n        # 从环境变量或参数获取配置\n        self.base_url = base_url or os.environ.get('LLM_API_BASE_URL', 'api.openai.com')\n        self.api_key = api_key or os.environ.get('LLM_API_KEY', '')\n        self.model = model or os.environ.get('LLM_MODEL', 'gpt-4o')\n        self.max_tokens = int(os.environ.get('LLM_MAX_TOKENS', '400'))\n        \n        # 解析base_url，分离主机名和路径\n        parsed_url = urlparse(f\"https://{self.base_url}\" if not self.base_url.startswith(('http://', 'https://')) else self.base_url)\n        self.host = parsed_url.netloc\n        self.base_path = parsed_url.path.rstrip('/')\n        \n        # 验证配置\n        if not self.api_key:\n            logger.warning(\"API密钥未设置，图像识别功能将不可用\")\n    \n    def recognize_image(self, image_path: str) -> Optional[str]:\n        \"\"\"\n        识别图像内容\n        \n        Args:\n            image_path: 图像文件路径\n            \n        Returns:\n            识别结果文本，如果识别失败则返回None\n        \"\"\"\n        if not self.api_key:\n            logger.error(\"API密钥未设置，无法进行图像识别\")\n            return None\n        \n        if not os.path.exists(image_path):\n            logger.error(f\"图像文件不存在: {image_path}\")\n            return None\n        \n        try:\n            # 检查是否是图像文件\n            if not image_path.lower().endswith(('.png', '.jpg', '.jpeg', '.gif', '.bmp', '.webp')):\n                logger.warning(f\"文件可能不是图像: {image_path}\")\n            \n            # 准备请求数据\n            messages = [\n                {\n                    \"role\": \"user\",\n                    \"content\": [\n                        {\n                            \"type\": \"text\",\n                            \"text\": \"请详细描述这张图片中的内容，包括图片中的文字、图形、颜色等元素。\"\n                        },\n                        {\n                            \"type\": \"image_url\",\n                            \"image_url\": {\n                                \"url\": self._get_image_url(image_path)\n                            }\n                        }\n                    ]\n                }\n            ]\n            \n            payload = json.dumps({\n                \"model\": self.model,\n                \"stream\": False,\n                \"messages\": messages,\n                \"max_tokens\": self.max_tokens\n            })\n            \n            headers = {\n                'Accept': 'application/json',\n                'Authorization': f'Bearer {self.api_key}',\n                'Content-Type': 'application/json'\n            }\n            \n            # 建立连接并发送请求\n            conn = http.client.HTTPSConnection(self.host)\n            conn.request(\"POST\", f\"{self.base_path}/v1/chat/completions\", payload, headers)\n            \n            # 获取响应\n            response = conn.getresponse()\n            data = response.read().decode(\"utf-8\")\n            conn.close()\n            \n            # 解析响应\n            response_data = json.loads(data)\n            \n            if 'error' in response_data:\n                logger.error(f\"API错误: {response_data['error']['message']}\")\n                return None\n            \n            # 提取识别结果\n            if 'choices' in response_data and len(response_data['choices']) > 0:\n                result = response_data['choices'][0]['message']['content']\n                return result\n            else:\n                logger.error(\"API响应中未找到识别结果\")\n                return None\n            \n        except Exception as e:\n            logger.error(f\"图像识别过程中发生错误: {str(e)}\")\n            return None\n    \n    def _get_image_url(self, image_path: str) -> str:\n        \"\"\"\n        获取图像URL，如果是本地文件则转换为base64编码\n        \n        Args:\n            image_path: 图像文件路径\n            \n        Returns:\n            图像URL或base64编码的数据URL\n        \"\"\"\n        # 如果是URL则直接返回\n        if image_path.startswith(('http://', 'https://')):\n            return image_path\n        \n        # 否则将本地文件转换为base64编码的数据URL\n        try:\n            with open(image_path, \"rb\") as image_file:\n                encoded_string = base64.b64encode(image_file.read()).decode('utf-8')\n            \n            # 确定MIME类型\n            mime_type = self._get_mime_type(image_path)\n            return f\"data:{mime_type};base64,{encoded_string}\"\n        except Exception as e:\n            logger.error(f\"图像文件读取或编码失败: {str(e)}\")\n            raise\n    \n    def _get_mime_type(self, file_path: str) -> str:\n        \"\"\"\n        根据文件扩展名获取MIME类型\n        \n        Args:\n            file_path: 文件路径\n            \n        Returns:\n            MIME类型字符串\n        \"\"\"\n        extension = os.path.splitext(file_path)[1].lower()\n        mime_types = {\n            '.jpg': 'image/jpeg',\n            '.jpeg': 'image/jpeg',\n            '.png': 'image/png',\n            '.gif': 'image/gif',\n            '.bmp': 'image/bmp',\n            '.webp': 'image/webp'\n        }\n        return mime_types.get(extension, 'application/octet-stream')\n\n\n# 创建默认客户端实例，方便直接导入使用\ndefault_client = ImageRecognitionClient()\n\ndef recognize_image(image_path: str) -> Optional[str]:\n    \"\"\"\n    使用默认客户端识别图像内容\n    \n    Args:\n        image_path: 图像文件路径\n        \n    Returns:\n        识别结果文本，如果识别失败则返回None\n    \"\"\"\n    return default_client.recognize_image(image_path) "
        }
    ]
}