debug: 修改错误标识可网络搜索模型的错误

server/adapters/glm_adapter.py

@@ -24,7 +24,7 @@ GLM_MODELS = [
         max_tokens=128000,
         provider="ZhipuAI",
         supports_thinking=True,
-        supports_web_search=True,
+        supports_web_search=False,
         supports_vision=True,
         supports_files=True,
     ),
@@ -46,7 +46,7 @@ GLM_MODELS = [
         max_tokens=128000,
         provider="ZhipuAI",
         supports_thinking=False,
-        supports_web_search=True,
+        supports_web_search=False,
         supports_vision=True,
         supports_files=True,
     ),

server/utils/test_glm_search.py

@@ -1,37 +1,84 @@
+"""
+GLM 适配器测试脚本
+测试 GLMAdapter 的流式和非流式调用，包括联网搜索功能
+"""
+
 import asyncio
 import os
 import sys
 from pathlib import Path
 
 # Add project root to sys.path
-root_dir = Path(__file__).parent
+root_dir = Path(__file__).parent.parent
 sys.path.insert(0, str(root_dir))
 
-# Set API key from .env if needed
 from dotenv import load_dotenv
 
-from utils.glm_adapter import _ensure_venv, glm_chat_sync, glm_stream_generator
+from adapters.glm_adapter import GLMAdapter
+from adapters.base import ChatCompletionRequest
 
 load_dotenv()
 
 
 async def test_stream():
-    msgs = [{"role": "user", "content": "今天北京天气怎样？"}]
+    """测试流式调用（联网搜索）"""
-    print("Testing stream...")
+    adapter = GLMAdapter()
-    async for chunk in glm_stream_generator(
+
-        msgs, "glm-4.5-air", 0.7, 1024, web_search=True
+    if not adapter.is_available():
-    ):
+        print("错误：未配置 ZHIPU_API_KEY 或 GLM_API_KEY")
+        return
+
+    request = ChatCompletionRequest(
+        model="glm-4.6v",
+        messages=[{"role": "user", "content": "今天北京天气怎样？"}],
+        stream=True,
+        temperature=0.7,
+        max_tokens=1024,
+        web_search=True,
+    )
+
+    print("Testing stream with web_search...")
+    response = await adapter.chat(request)
+
+    # 流式响应是 StreamingResponse，需要手动读取
+    async for chunk in response.body_iterator:
+        # body_iterator 已经返回字符串
         print(chunk, end="")
 
 
-def test_sync():
+async def test_sync():
-    msgs = [{"role": "user", "content": "今天几号？武汉天气怎样？"}]
+    """测试非流式调用（联网搜索）"""
-    print("Testing sync...")
+    adapter = GLMAdapter()
-    res = glm_chat_sync(msgs, "glm-4.5-air", 0.7, 1024, web_search=True)
+
-    print(res)
+    if not adapter.is_available():
+        print("错误：未配置 ZHIPU_API_KEY 或 GLM_API_KEY")
+        return
+
+    request = ChatCompletionRequest(
+        model="glm-4-flash",
+        messages=[{"role": "user", "content": "今天几号？武汉天气怎样？"}],
+        stream=False,
+        temperature=0.7,
+        max_tokens=1024,
+        web_search=True,
+    )
+
+    print("Testing sync with web_search...")
+    response = await adapter.chat(request)
+
+    # 非流式响应返回 JSONResponse
+    if hasattr(response, "body"):
+        import json
+
+        data = json.loads(response.body)
+        content = data.get("choices", [{}])[0].get("message", {}).get("content", "")
+        print(f"Response: {content}")
+    else:
+        print(f"Response: {response}")
 
 
 if __name__ == "__main__":
-    _ensure_venv()
+    # 运行流式测试
-    # test_sync()
+    # asyncio.run(test_stream())
-    asyncio.run(test_stream())
+    # 运行非流式测试
+    asyncio.run(test_sync())