Spaces:
Sleeping
Sleeping
File size: 17,122 Bytes
4fc29a0 b3e6518 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e88b429 4fc29a0 e3d6247 4fc29a0 e3d6247 e88b429 4fc29a0 e3d6247 e88b429 e3d6247 e88b429 e3d6247 e88b429 e3d6247 b3e6518 e3d6247 9cd681c b3e6518 9cd681c b4a3e27 9cd681c b4a3e27 b3e6518 9cd681c b3e6518 9cd681c b3e6518 9cd681c b3e6518 9cd681c b3e6518 9cd681c b3e6518 9cd681c b3e6518 9cd681c b3e6518 9cd681c b3e6518 e3d6247 e88b429 b4a3e27 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 bb95a94 b3e6518 e88b429 b3e6518 0fd5c22 bb95a94 0fd5c22 bb95a94 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 4fc29a0 e3d6247 0fd5c22 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 | # 导入所需的库
import json # 用于JSON数据处理
import time # 用于时间相关操作
import traceback # 用于异常堆栈跟踪
import uuid
from datetime import datetime
from typing import Optional, Dict, Any
from fastapi import FastAPI, HTTPException, Request, Depends, Response # FastAPI框架相关组件
from fastapi.middleware.cors import CORSMiddleware # CORS中间件
from fastapi.responses import StreamingResponse, JSONResponse # 特殊响应类型
import httpx # 异步HTTP客户端
import logging # 日志管理
import random # 随机数生成
import uvicorn # ASGI服务器
import asyncio # 异步IO
# API端点配置
QWEN_API_URL = 'https://chat.qwenlm.ai/api/chat/completions' # 通达API聊天完成接口
QWEN_MODELS_URL = 'https://chat.qwenlm.ai/api/models' # 模型列表接口
MAX_RETRIES = 3 # 最大重试次数
RETRY_DELAY = 1 # 重试延迟时间(秒)
# 配置日志
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s - %(levelname)s - %(message)s",
)
logger = logging.getLogger(__name__)
# 创建FastAPI应用实例
app = FastAPI()
# 添加CORS中间件
app.add_middleware(
CORSMiddleware,
allow_origins=["*"], # 允许所有来源
allow_credentials=True, # 允许携带凭证
allow_methods=["*"], # 允许所有HTTP方法
allow_headers=["*"], # 允许所有请求头
)
# 创建异步HTTP客户端
client = httpx.AsyncClient()
async def fetch_with_retry(url, options, retries=MAX_RETRIES):
"""
带重试机制的异步请求函数
参数:
url: 请求URL
options: 请求选项
retries: 重试次数
"""
last_error = None
auth_header = options.get('headers', {}).get('Authorization', '')
if auth_header.startswith('Bearer '):
logger.info(f"Processing request with session identifier: {auth_header[7:]}")
for i in range(retries):
try:
async with httpx.AsyncClient() as client:
# 设置请求头
if 'headers' in options:
client.headers.update(options['headers'])
# 发送请求
response = await client.request(
method=options.get('method', 'GET'),
url=url,
content=options.get('body'),
timeout=60
)
# 检查响应类型和状态码
content_type = response.headers.get('content-type', '')
if 'text/html' in content_type or response.status_code == 500:
last_error = {
'status': response.status_code,
'contentType': content_type,
'responseText': response.text[:1000],
'headers': dict(response.headers)
}
if i < retries - 1:
logger.error(f"Retry attempt {i+1} for session {auth_header[7:]} failed")
await asyncio.sleep(RETRY_DELAY * (i + 1))
continue
return response
except Exception as error:
last_error = error
logger.error(f"Connection error for session {auth_header[7:]} on attempt {i+1}")
traceback.print_exc()
if i < retries - 1:
await asyncio.sleep(RETRY_DELAY * (i + 1))
continue
# 所有重试都失败后抛出异常
raise Exception(json.dumps({
'error': True,
'message': 'All retry attempts failed',
'lastError': str(last_error),
'retries': retries
}))
async def process_line(line, previous_content):
"""
处理流式响应中的单行数据,避免内容重复
"""
try:
data = json.loads(line[6:]) # 解析JSON数据
if data.get('choices') and data['choices'][0].get('delta') and data['choices'][0]['delta'].get('content'):
current_content = data['choices'][0]['delta']['content']
new_content = current_content
# 避免内容重复
if current_content.startswith(previous_content) and len(previous_content) > 0:
new_content = current_content[len(previous_content):]
# 只有在有新内容时才构建新的响应数据
if new_content:
new_data = {
'choices': [{
'delta': {
'content': new_content
}
}]
}
return f"data: {json.dumps(new_data)}\n\n", new_content
return None, None
else:
return f"data: {json.dumps(data)}\n\n", None
except Exception:
return f"{line}\n\n", None
# async def handle_stream(response, previous_content):
# """
# 处理流式响应,确保实时发送到客户端
# """
# buffer = ''
# async for chunk in response.aiter_lines(encoding='utf-8'):
# try:
# # decoded_chunk = chunk.decode('utf-8')
# decoded_chunk = chunk.strip()
# print('decoded_chunk: ', decoded_chunk, ' ::end')
# buffer += decoded_chunk
#
# # 立即处理每个完整的数据行
# while '\n' in buffer:
# # print('buffer: ', buffer)
# line, buffer = buffer.split('\n', 1)
# line = line.strip()
# if line.startswith('data: '):
# processed_line, new_content = await process_line(line, previous_content)
# if processed_line:
# # 确保每个响应都以正确的SSE格式发送
# # print('11data: ', json.dumps(processed_line, indent=4))
# yield processed_line
# # 立即刷新输出
# await asyncio.sleep(0)
# if new_content:
# previous_content = previous_content + new_content
#
# # print("data: [DONE]")
# # yield "data: [DONE]\n\n"
#
# except Exception as e:
# logger.error(f"Error processing chunk: {str(e)}")
# yield f"data: {{\"error\":true,\"message\":\"{str(e)}\"}}\n\n"
# continue
#
# # 处理剩余的buffer
# if buffer.strip():
# line = buffer.strip()
# if line.startswith('data: '):
# processed_line, new_content = await process_line(line, previous_content)
# if processed_line:
# yield processed_line
#
# print("data: [DONE]")
# yield "data: [DONE]\n\n"
def create_chat_completion_data(
content: str, model: str, timestamp: int, finish_reason: Optional[str] = None
) -> Dict[str, Any]:
return {
"id": f"chatcmpl-{uuid.uuid4()}",
"object": "chat.completion.chunk",
"created": timestamp,
"model": model,
"choices": [
{
"index": 0,
"delta": {"content": content, "role": "assistant"},
"finish_reason": finish_reason,
}
],
"usage": None,
}
async def handle_stream(response, previous_content):
"""
优化后的流式响应处理函数
主要改进:
1. 使用简单的字符串切片而不是startswith比较
2. 增加内容缓存来减少重复处理
3. 优化内存使用
"""
timestamp = int(datetime.now().timestamp())
content_cache = "" # 用于缓存已处理的内容
async def send_chunk(content: str):
"""内部函数:格式化并发送数据块"""
chunk_data = {
"id": f"chatcmpl-{uuid.uuid4()}",
"object": "chat.completion.chunk",
"created": timestamp,
"model": "qwen",
"choices": [{
"index": 0,
"delta": {
"content": content,
"role": "assistant"
},
"finish_reason": None
}]
}
return f"data: {json.dumps(chunk_data)}\n\n"
async for chunk in response.aiter_lines():
if not chunk or not chunk.startswith('data: '):
continue
try:
# 解析数据
data = json.loads(chunk[6:]) # 去掉 'data: ' 前缀
# 提取新内容
if not (data.get('choices') and
data['choices'][0].get('delta') and
data['choices'][0]['delta'].get('content')):
continue
current_content = data['choices'][0]['delta']['content']
# 智能差异检测
if content_cache:
# 找到新内容的起始位置
for i in range(min(len(current_content), len(content_cache))):
if current_content[i] != content_cache[i]:
new_content = current_content[i:]
break
else:
# 如果前面的内容都相同,新内容就是超出的部分
new_content = current_content[len(content_cache):]
else:
new_content = current_content
# 只有真正有新内容时才发送
if new_content:
yield await send_chunk(new_content)
content_cache = current_content # 更新缓存
except json.JSONDecodeError as e:
logger.error(f"JSON解析错误: {str(e)}")
continue
except Exception as e:
logger.error(f"处理数据流时发生错误: {str(e)}")
yield f"data: {{\"error\":true,\"message\":\"{str(e)}\"}}\n\n"
# 发送结束标记
yield "data: [DONE]\n\n"
async def get_openai_auth_headers(request: Request) -> dict:
"""
获取认证头信息
参数:
request: 请求对象
返回:
包含认证信息的字典
"""
auth_header = request.headers.get("Authorization")
if not auth_header:
raise HTTPException(status_code=401, detail="Missing Authorization header")
logger.info(f"New request authenticated with session {auth_header[7:]}")
return {"Authorization": auth_header}
async def make_request(method, url, headers, body, api_keys=None, retry_count=0):
"""
发送请求的通用函数,支持多个API密钥
"""
try:
if api_keys and len(api_keys) > 1:
# 多个API密钥的情况
remaining_keys = api_keys.copy()
while remaining_keys and retry_count < 3:
selected_key = random.choice(remaining_keys)
remaining_keys.remove(selected_key)
headers = {**headers, "Authorization": f"Bearer {selected_key}"}
logger.info(f"Attempting request with API key: {selected_key}")
try:
async with httpx.AsyncClient() as client:
r = await client.request(
method,
url,
headers=headers,
content=body,
timeout=600
)
if r.status_code < 400:
return r
logger.error(f"Request failed with key {selected_key}, status code: {r.status_code}")
except Exception as e:
logger.error(f"Request failed with key {selected_key}: {str(e)}")
retry_count += 1
raise HTTPException(status_code=500, detail="All API keys failed")
else:
# 单个API密钥的情况
while retry_count < 3:
single_key = api_keys[0] if api_keys else headers.get("authorization", "").replace("Bearer ", "").strip()
request_headers = {**headers, "Authorization": f"Bearer {single_key}"}
logger.info(f"Attempting request with API key: {single_key}")
try:
async with httpx.AsyncClient() as client:
r = await client.request(
method,
url,
headers=request_headers,
content=body,
timeout=600
)
if r.status_code < 400:
return r
logger.error(f"Request attempt {retry_count + 1} failed for session {single_key}")
except Exception as e:
logger.error(f"Connection attempt {retry_count + 1} failed for session {single_key}")
retry_count += 1
raise HTTPException(status_code=500, detail="Request failed after 3 retries")
except Exception as e:
logger.error(f"Request failed: {str(e)}")
raise HTTPException(status_code=500, detail=str(e))
# API路由定义
@app.api_route(
"/api/chat/completions",
methods=["POST", "OPTIONS"],
)
async def chat_completions(request: Request, auth_headers: dict = Depends(get_openai_auth_headers)):
"""
处理聊天完成请求的端点
"""
# 处理请求头
headers = dict(request.headers)
if "content-length" in headers:
del headers["content-length"]
if "host" in headers:
del headers["host"]
# 获取请求体
request_body = await request.body()
try:
request_data = json.loads(request_body.decode('utf-8'))
except json.JSONDecodeError:
raise HTTPException(status_code=400, detail="Invalid JSON")
# 提取请求参数
messages = request_data.get('messages')
stream = request_data.get('stream', False)
model = request_data.get('model')
max_tokens = request_data.get('max_tokens')
if not model:
raise HTTPException(status_code=400, detail="Model parameter is required")
# 构建请求数据
qwen_request = {
'model': model,
'messages': messages,
'stream': stream
}
if max_tokens is not None:
qwen_request['max_tokens'] = max_tokens
try:
# 发送请求到Qwen API
response = await fetch_with_retry(QWEN_API_URL, {
'method': 'POST',
'headers': {
'Content-Type': 'application/json',
**auth_headers
},
'body': json.dumps(qwen_request),
'stream': stream
})
# 处理响应
if stream:
previous_content = ''
return StreamingResponse(handle_stream(response, previous_content), media_type="text/event-stream", headers={
'Cache-Control': 'no-cache, no-transform',
'Connection': 'keep-alive',
'X-Accel-Buffering': 'no', # 禁用 Nginx 缓冲
'Content-Type': 'text/event-stream',
'Transfer-Encoding': 'chunked'
},
status_code=200,
background=None # 确保不会在后台处理
)
else:
return Response(content=response.content, status_code=response.status_code, headers=response.headers)
except Exception as error:
raise HTTPException(status_code=500, detail=str(error))
@app.get("/api/models")
async def models(request: Request, auth_headers: dict = Depends(get_openai_auth_headers)):
"""获取可用模型列表的端点"""
try:
response = await fetch_with_retry(QWEN_MODELS_URL, {
'method': 'GET',
'headers': {
'Content-Type': 'application/json',
**auth_headers
},
'timeout': 30
})
response_data = response.json()
return JSONResponse(content=response_data)
except Exception as e:
logger.error(f"Error in /api/models: {str(e)}")
raise HTTPException(status_code=500, detail=str(e))
@app.get('/')
async def index(request: Request):
"""根路径重定向"""
return Response(status_code=302, headers={"Location": "https://chat.qwenlm.ai"})
@app.get('/{path:path}')
@app.post('/{path:path}')
async def redirect_all(path: str, request: Request):
"""处理所有其他路径的重定向"""
# 检查是否包含敏感关键词
if any(keyword in path.lower() for keyword in ['php', 'admin', 'login', 'wp-admin', 'manager', 'user', 'signin']):
return Response(status_code=301, headers={"Location": "http://127.0.0.1"})
if request.method == 'POST':
return Response(status_code=301, headers={"Location": "http://127.0.0.1"})
return Response(status_code=302, headers={"Location": "https://linux.do/u/f-droid"})
# 主程序入口
if __name__ == "__main__":
uvicorn.run(
app,
host="0.0.0.0",
port=8080,
loop="asyncio",
timeout_keep_alive=65,
access_log=True,
log_level="debug",
http="h11",
limit_concurrency=1000,
backlog=2048
)
|