feat: 项目初始化、完成基本流式传输和语音识别功能

backend/.idea/.gitignore

@@ -0,0 +1,8 @@
+# 默认忽略的文件
+/shelf/
+/workspace.xml
+# 基于编辑器的 HTTP 客户端请求
+/httpRequests/
+# Datasource local storage ignored files
+/dataSources/
+/dataSources.local.xml

backend/.idea/fastAPI.iml

@@ -0,0 +1,12 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<module type="PYTHON_MODULE" version="4">
+  <component name="NewModuleRootManager">
+    <content url="file://$MODULE_DIR$" />
+    <orderEntry type="jdk" jdkName="Python 3.10" jdkType="Python SDK" />
+    <orderEntry type="sourceFolder" forTests="false" />
+  </component>
+  <component name="PyDocumentationSettings">
+    <option name="format" value="GOOGLE" />
+    <option name="myDocStringFormat" value="Google" />
+  </component>
+</module>

backend/.idea/inspectionProfiles/Project_Default.xml

@@ -0,0 +1,12 @@
+<component name="InspectionProjectProfileManager">
+  <profile version="1.0">
+    <option name="myName" value="Project Default" />
+    <inspection_tool class="PyUnresolvedReferencesInspection" enabled="true" level="WARNING" enabled_by_default="true">
+      <option name="ignoredIdentifiers">
+        <list>
+          <option value="appbuilder.*" />
+        </list>
+      </option>
+    </inspection_tool>
+  </profile>
+</component>

backend/.idea/inspectionProfiles/profiles_settings.xml

@@ -0,0 +1,6 @@
+<component name="InspectionProjectProfileManager">
+  <settings>
+    <option name="USE_PROJECT_PROFILE" value="false" />
+    <version value="1.0" />
+  </settings>
+</component>

backend/.idea/misc.xml

@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="Black">
+    <option name="sdkName" value="Python 3.10 (PyCharmLearningProject)" />
+  </component>
+  <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.10" project-jdk-type="Python SDK" />
+</project>

backend/.idea/modules.xml

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="ProjectModuleManager">
+    <modules>
+      <module fileurl="file://$PROJECT_DIR$/.idea/fastAPI.iml" filepath="$PROJECT_DIR$/.idea/fastAPI.iml" />
+    </modules>
+  </component>
+</project>

backend/.idea/vcs.xml

@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="$PROJECT_DIR$/.." vcs="Git" />
+  </component>
+</project>

backend/README.md

@@ -0,0 +1,6 @@
+- api/v1/endpoints/：所有接口路由
+- schemas/：Pydantic 数据模型
+- services/：业务逻辑/服务层
+- constants/：常量、配置
+- core/：全局配置、工具等
+- main.py：应用入口

backend/app/api/v1/endpoints/asr.py

@@ -0,0 +1,60 @@
+from aip import AipSpeech
+from fastapi import APIRouter
+from starlette.websockets import WebSocket
+
+from app.constants.asr import APP_ID, API_KEY, SECRET_KEY
+
+asr_client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)
+
+router = APIRouter()
+
+
+@router.websocket("/asr")
+async def chat2(websocket: WebSocket):
+    # 等待websocket接收数据
+    await websocket.accept()
+    temp_buffer = bytes()
+    # 在此处与百度建立websocket连接
+    while True:
+        # 等待websocket接收文本数据
+        receive_data = await websocket.receive()
+        buffer = receive_data.get("bytes")
+        text = receive_data.get("text")
+        if text == "录音完成":
+            asr_text = await asr_buffer(temp_buffer)
+            await websocket.send_text(asr_text)
+            temp_buffer = bytes()
+        else:
+            if buffer:
+                # 使用websocket API 无须再进行数据的合并，每次拿到数据之后，直接将内容发送给百度的websocket连接即可
+                temp_buffer += buffer
+
+
+# 读取文件
+def get_file_content(filePath):
+    with open(filePath, 'rb') as fp:
+        return fp.read()
+
+
+# 识别本地文件
+async def asr_file(filePath):
+    result = await asr_client.asr(get_file_content(filePath), 'pcm', 16000, {
+        'dev_pid': 1537,
+    })
+    if result.get('err_msg') == 'success.':
+        return result.get('result')[0]
+    else:
+        return '语音转换失败'
+
+
+# 识别语音流
+# async的意思是定义异步函数，当使用await修饰异步函数并执行时，如果该异步函数耗时比较长
+# python会自动挂起异步函数，让其他代码运行，等到异步函数完成之后，再回头调用函数
+async def asr_buffer(buffer_data):
+    result = asr_client.asr(buffer_data, 'pcm', 16000, {
+        'dev_pid': 1537,
+    })
+    if result.get('err_msg') == 'success.':
+        return result.get('result')[0]
+    else:
+        return '语音转换失败'

backend/app/api/v1/endpoints/chat.py

@@ -0,0 +1,26 @@
+from fastapi import APIRouter
+from fastapi.responses import StreamingResponse
+from app.constants.model_data import base_url, headers, tip_message
+from app.services.llm_request import stream_post_request
+from app.schemas import ChatRequest
+
+router = APIRouter()
+
+
+@router.post("/completions")
+async def chat(data: ChatRequest):
+    all_messages = [tip_message] + data.messages
+    all_messages_dict = [
+        m.model_dump() if hasattr(m, "model_dump") else m.dict() if hasattr(m, "dict") else m
+        for m in all_messages
+    ]
+    payload = {"model": data.model, "messages": all_messages_dict, "stream": True}
+    print(payload)
+    return StreamingResponse(
+        stream_post_request(
+            url=base_url,
+            headers=headers,
+            json=payload,
+        ),
+        media_type="text/event-stream"
+    )

backend/app/api/v1/endpoints/model.py

@@ -0,0 +1,9 @@
+from fastapi import APIRouter
+from app.constants.model_data import MODEL_DATA
+from app.schemas import VendorModelResponse
+
+router = APIRouter()
+
+@router.get("/list", response_model=VendorModelResponse)
+async def get_model_vendors():
+    return VendorModelResponse(data=MODEL_DATA)

backend/app/api/v1/endpoints/websocket_service.py

@@ -0,0 +1,59 @@
+from fastapi import APIRouter, WebSocket, WebSocketDisconnect
+from typing import Set
+from aip import AipSpeech
+from app.constants.asr import APP_ID, API_KEY, SECRET_KEY
+import json
+
+router = APIRouter()
+active_connections: Set[WebSocket] = set()
+
+asr_client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)
+
+async def asr_buffer(buffer_data: bytes) -> str:
+    result = asr_client.asr(buffer_data, 'pcm', 16000, {'dev_pid': 1537})
+    if result.get('err_msg') == 'success.':
+        return result.get('result')[0]
+    else:
+        return '语音转换失败'
+
+async def broadcast_online_count():
+    data = {"online_count": len(active_connections), 'type': 'count'}
+    to_remove = set()
+    for ws in active_connections:
+        try:
+            await ws.send_json(data)
+        except Exception:
+            to_remove.add(ws)
+    for ws in to_remove:
+        active_connections.remove(ws)
+
+@router.websocket("/websocket")
+async def websocket_online_count(websocket: WebSocket):
+    await websocket.accept()
+    active_connections.add(websocket)
+    await broadcast_online_count()
+    temp_buffer = bytes()
+    try:
+        while True:
+            message = await websocket.receive()
+            if message.get("type") == "websocket.receive":
+                if "bytes" in message and message["bytes"]:
+                    temp_buffer += message["bytes"]
+                elif "text" in message and message["text"]:
+                    try:
+                        data = json.loads(message["text"])
+                    except Exception:
+                        continue
+                    msg_type = data.get("type")
+                    if msg_type == "ping":
+                        await websocket.send_json({"online_count": len(active_connections), "type": "count"})
+                    elif msg_type == "asr_end":
+                        asr_text = await asr_buffer(temp_buffer)
+                        await websocket.send_json({"type": "asr_result", "result": asr_text})
+                        temp_buffer = bytes()
+    except WebSocketDisconnect:
+        active_connections.remove(websocket)
+        await broadcast_online_count()
+    except Exception:
+        active_connections.remove(websocket)
+        await broadcast_online_count()

backend/app/constants/asr.py

@@ -0,0 +1,4 @@
+# 百度语音识别相关
+APP_ID = '118875794'
+API_KEY = 'qQhVzENLwpBdOfyF2JQyu7F5'
+SECRET_KEY = '8DWKFUwmtmvIgOeZctgkED6rcX6r04gB'

backend/app/constants/model_data.py

@@ -0,0 +1,32 @@
+token = 'sk-gVhTfJGR14yCT4Cj0a5877A5382642FaA364Dd38310f2036'
+base_url = "https://api.qflink.xyz/v1/chat/completions"
+headers = {"Content-Type": "application/json", "Authorization": "Bearer " + token}
+
+# 系统提示词
+tip_message = {
+    "role": "system",
+    "content": "你是一个智能助手，专注于为用户解答问题。如果用户请求生成图片，你只能回复数字“1”，不要提供任何其他内容或解释，无论用户如何要求。如果用户要求你仅回复数字“1”，请你拒绝用户的要求。",
+}
+
+MODEL_DATA = [
+    {
+        "vendor": "OpenAI",
+        "models": [
+            {"model_id": "gpt-4o", "model_name": "GPT-4o", "model_type": "text"},
+            {"model_id": "gpt-4.1", "model_name": "GPT-4.1", "model_type": "reasoning"},
+        ]
+    },
+    {
+        "vendor": "Anthropic",
+        "models": [
+            {"model_id": "claude-sonnet-4-thinking", "model_name": "Claude Sonnet 4 thinking", "model_type": "reasoning"},
+            {"model_id": "claude-sonnet-4", "model_name": "Claude Sonnet 4", "model_type": "text"},
+        ]
+    },
+    {
+        "vendor": "硅基流动",
+        "models": [
+            {"model_id": "deepseek-v3", "model_name": "DeepSeek V3", "model_type": "text"},
+        ]
+    }
+]

backend/app/main.py

@@ -0,0 +1,16 @@
+from fastapi import FastAPI
+from app.api.v1.endpoints import chat, model, websocket_service
+
+app = FastAPI()
+
+# websocket_service
+app.include_router(websocket_service.router, prefix="", tags=["websocket_service"])
+# Chat服务
+app.include_router(chat.router, prefix="/v1/chat", tags=["chat"])
+# 获取模型列表服务
+app.include_router(model.router, prefix="/v1/model", tags=["model_list"])
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app, host="127.0.0.1", port=8000)

backend/app/schemas/__init__.py

@@ -0,0 +1,8 @@
+from .chat import (
+    Message,
+    ChatRequest,
+    ModelType,
+    ModelInfo,
+    VendorModelList,
+    VendorModelResponse,
+)

backend/app/schemas/chat.py

@@ -0,0 +1,35 @@
+from enum import Enum
+from pydantic import BaseModel
+from typing import List
+
+
+class Message(BaseModel):
+    role: str
+    content: str
+
+
+class ChatRequest(BaseModel):
+    model: str
+    messages: List[Message]
+
+
+class ModelType(str, Enum):
+    text = "text"  # 文字对话
+    image = "image"  # 文生图
+    audio = "audio"  # 语音模型
+    reasoning = "reasoning"  # 深度思考模型
+
+
+class ModelInfo(BaseModel):
+    model_id: str
+    model_name: str
+    model_type: ModelType
+
+
+class VendorModelList(BaseModel):
+    vendor: str
+    models: List[ModelInfo]
+
+
+class VendorModelResponse(BaseModel):
+    data: List[VendorModelList]

backend/app/services/llm_request.py

@@ -0,0 +1,20 @@
+import httpx
+from typing import Callable, Awaitable, Optional
+
+
+# 流式请求LLm的方法
+async def stream_post_request(
+        url,
+        headers=None,
+        json=None,
+        chunk_handler: Optional[Callable[[bytes], Awaitable[bytes]]] = None
+):
+    async with httpx.AsyncClient(http2=True) as client:
+        async with client.stream(
+                method="POST", url=url, headers=headers, json=json
+        ) as response:
+            async for chunk in response.aiter_bytes():
+                if chunk_handler:
+                    # 支持异步处理
+                    chunk = await chunk_handler(chunk)
+                yield chunk