Update app.py
Browse files
app.py
CHANGED
@@ -1,298 +1,361 @@
|
|
1 |
-
import os
|
2 |
-
import
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
from fastapi
|
7 |
-
import
|
8 |
-
from
|
9 |
-
import
|
10 |
-
|
11 |
-
import
|
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 |
-
logger.info("HTTPX client
|
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 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import time
|
3 |
+
|
4 |
+
import httpx
|
5 |
+
import json
|
6 |
+
from fastapi import FastAPI, Request, HTTPException, Response, Depends
|
7 |
+
from fastapi.security import APIKeyHeader
|
8 |
+
from fastapi.responses import StreamingResponse, JSONResponse
|
9 |
+
import logging
|
10 |
+
from contextlib import asynccontextmanager
|
11 |
+
import typing
|
12 |
+
import itertools # For key rotation
|
13 |
+
from typing import List, Optional, Dict, Any, Union
|
14 |
+
|
15 |
+
from trio import sleep
|
16 |
+
|
17 |
+
# --- Configuration ---
|
18 |
+
|
19 |
+
# --- Client Authentication (Proxy Access) ---
|
20 |
+
# Load Allowed Client API Keys (for clients talking to this proxy)
|
21 |
+
ALLOWED_API_KEYS_STR = os.getenv("ALLOWED_API_KEYS")
|
22 |
+
if not ALLOWED_API_KEYS_STR:
|
23 |
+
raise ValueError("REQUIRED: ALLOWED_API_KEYS environment variable (comma-separated keys for clients) not set.")
|
24 |
+
ALLOWED_KEYS = set(key.strip() for key in ALLOWED_API_KEYS_STR.split(',') if key.strip())
|
25 |
+
if not ALLOWED_KEYS:
|
26 |
+
raise ValueError("ALLOWED_API_KEYS must contain at least one non-empty key.")
|
27 |
+
logging.info(f"Loaded {len(ALLOWED_KEYS)} allowed client API keys.")
|
28 |
+
|
29 |
+
# --- Upstream API Configuration ---
|
30 |
+
# URL to fetch upstream API keys from (one key per line)
|
31 |
+
UPSTREAM_KEYS_URL = os.getenv("UPSTREAM_KEYS_URL")
|
32 |
+
|
33 |
+
# Optional: A single fallback/default upstream key (used if URL fetch fails or isn't provided)
|
34 |
+
# Or required if the upstream target needs a key in a different way sometimes.
|
35 |
+
# Let's make it optional now.
|
36 |
+
DEFAULT_OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
|
37 |
+
|
38 |
+
# Upstream API Base URL
|
39 |
+
OPENAI_API_BASE = os.getenv("OPENAI_API_BASE", "https://models.aixplain.com/api/v1")
|
40 |
+
OPENAI_CHAT_ENDPOINT = f"{OPENAI_API_BASE.rstrip('/')}/chat/completions"
|
41 |
+
|
42 |
+
if not UPSTREAM_KEYS_URL and not DEFAULT_OPENAI_API_KEY:
|
43 |
+
raise ValueError("REQUIRED: Either UPSTREAM_KEYS_URL or OPENAI_API_KEY environment variable must be set for upstream authentication.")
|
44 |
+
|
45 |
+
# --- Logging ---
|
46 |
+
logging.basicConfig(level=logging.INFO)
|
47 |
+
logger = logging.getLogger(__name__)
|
48 |
+
|
49 |
+
# --- Authentication Dependency (Client -> Proxy) ---
|
50 |
+
api_key_header_auth = APIKeyHeader(name="Authorization", auto_error=False)
|
51 |
+
|
52 |
+
|
53 |
+
# OpenAI API请求模型
|
54 |
+
class ChatMessage:
|
55 |
+
role: str
|
56 |
+
content: str
|
57 |
+
|
58 |
+
async def verify_api_key(api_key_header: typing.Optional[str] = Depends(api_key_header_auth)):
|
59 |
+
"""Dependency to verify the client's API key provided to this proxy."""
|
60 |
+
if not api_key_header:
|
61 |
+
logger.warning("Missing Authorization header from client")
|
62 |
+
raise HTTPException(status_code=401, detail="Missing Authorization header")
|
63 |
+
parts = api_key_header.split()
|
64 |
+
if len(parts) != 2 or parts[0].lower() != "bearer":
|
65 |
+
logger.warning(f"Invalid Authorization header format from client.")
|
66 |
+
raise HTTPException(status_code=401, detail="Invalid Authorization header format. Use 'Bearer YOUR_KEY'.")
|
67 |
+
client_api_key = parts[1]
|
68 |
+
if client_api_key not in ALLOWED_KEYS:
|
69 |
+
truncated_key = client_api_key[:4] + "..." + client_api_key[-4:] if len(client_api_key) > 8 else client_api_key
|
70 |
+
logger.warning(f"Invalid Client API Key received: {truncated_key}")
|
71 |
+
raise HTTPException(status_code=403, detail="Invalid API Key provided")
|
72 |
+
logger.info(f"Client authenticated successfully (Key ending: ...{client_api_key[-4:]})")
|
73 |
+
return client_api_key
|
74 |
+
|
75 |
+
# --- Key Fetching and Rotation Logic ---
|
76 |
+
async def fetch_upstream_keys(url: str) -> list[str]:
|
77 |
+
"""Fetches keys from the given URL, one key per line."""
|
78 |
+
keys = []
|
79 |
+
try:
|
80 |
+
async with httpx.AsyncClient(timeout=15.0) as client: # Use a temporary client
|
81 |
+
logger.info(f"Fetching upstream API keys from: {url}")
|
82 |
+
response = await client.get(url)
|
83 |
+
response.raise_for_status() # Raise exception for 4xx/5xx status codes
|
84 |
+
content = response.text
|
85 |
+
keys = [line.strip() for line in content.splitlines() if line.strip()]
|
86 |
+
logger.info(f"Successfully fetched {len(keys)} upstream API keys.")
|
87 |
+
if not keys:
|
88 |
+
logger.warning(f"No valid keys found at {url}. The response was empty or contained only whitespace.")
|
89 |
+
return keys
|
90 |
+
except httpx.RequestError as e:
|
91 |
+
logger.error(f"Error fetching upstream keys from {url}: {e}")
|
92 |
+
return [] # Return empty list on fetch error
|
93 |
+
except httpx.HTTPStatusError as e:
|
94 |
+
logger.error(f"Error fetching upstream keys from {url}: Status {e.response.status_code}")
|
95 |
+
logger.error(f"Response body: {e.response.text}")
|
96 |
+
return [] # Return empty list on bad status
|
97 |
+
|
98 |
+
# --- HTTP Client and Key Iterator Management (Lifespan) ---
|
99 |
+
@asynccontextmanager
|
100 |
+
async def lifespan(app: FastAPI):
|
101 |
+
# --- Initialize Upstream Key Iterator ---
|
102 |
+
upstream_keys = []
|
103 |
+
if UPSTREAM_KEYS_URL:
|
104 |
+
upstream_keys = await fetch_upstream_keys(UPSTREAM_KEYS_URL)
|
105 |
+
|
106 |
+
if not upstream_keys:
|
107 |
+
logger.warning("No upstream keys fetched from URL or URL not provided.")
|
108 |
+
if DEFAULT_OPENAI_API_KEY:
|
109 |
+
logger.info("Using fallback OPENAI_API_KEY for upstream authentication.")
|
110 |
+
upstream_keys = [DEFAULT_OPENAI_API_KEY]
|
111 |
+
else:
|
112 |
+
# Critical failure - no keys available
|
113 |
+
logger.critical("FATAL: No upstream API keys available (URL fetch failed/empty and no fallback OPENAI_API_KEY). Exiting.")
|
114 |
+
# In a real scenario, you might want a more graceful shutdown or retry mechanism
|
115 |
+
# For simplicity here, we'll let it proceed but log critically. The requests will likely fail later.
|
116 |
+
# Or raise an exception here to prevent startup:
|
117 |
+
raise RuntimeError("Failed to load any upstream API keys. Cannot start service.")
|
118 |
+
|
119 |
+
# Store keys and create the cycling iterator in app.state
|
120 |
+
app.state.upstream_api_keys = upstream_keys
|
121 |
+
app.state.key_iterator = itertools.cycle(upstream_keys)
|
122 |
+
logger.info(f"Initialized key rotation with {len(upstream_keys)} keys.")
|
123 |
+
|
124 |
+
# --- Initialize HTTPX Client ---
|
125 |
+
logger.info("Initializing main HTTPX client...")
|
126 |
+
timeout = httpx.Timeout(5.0, read=180.0, write=5.0, connect=5.0)
|
127 |
+
client = httpx.AsyncClient(timeout=timeout) # No base_url needed if using full URLs
|
128 |
+
app.state.http_client = client # Store client in app.state
|
129 |
+
logger.info("HTTPX client initialized.")
|
130 |
+
|
131 |
+
yield # Application runs here
|
132 |
+
|
133 |
+
# --- Cleanup ---
|
134 |
+
logger.info("Closing HTTPX client...")
|
135 |
+
await app.state.http_client.aclose()
|
136 |
+
logger.info("HTTPX client closed.")
|
137 |
+
app.state.upstream_api_keys = [] # Clear keys
|
138 |
+
app.state.key_iterator = None
|
139 |
+
logger.info("Upstream keys cleared.")
|
140 |
+
|
141 |
+
# --- FastAPI App ---
|
142 |
+
app = FastAPI(lifespan=lifespan)
|
143 |
+
|
144 |
+
# --- Streaming Helper ---
|
145 |
+
async def yield_openai_chunks(response_body):
|
146 |
+
"""Asynchronously yields chunks from the upstream response stream."""
|
147 |
+
# (Content remains the same as before)
|
148 |
+
logger.info("Starting to stream chunks from upstream...")
|
149 |
+
try:
|
150 |
+
resp_json = json.loads(response_body)
|
151 |
+
|
152 |
+
for choices in resp_json["choices"]:
|
153 |
+
if "message" in choices:
|
154 |
+
choices["delta"] = choices["message"]
|
155 |
+
del choices["message"]
|
156 |
+
|
157 |
+
yield "data:" + json.dumps(resp_json) + "\n\n"
|
158 |
+
yield "data: [DONE]"
|
159 |
+
except Exception as e:
|
160 |
+
logger.error(f"Error during streaming upstream response: {e}")
|
161 |
+
finally:
|
162 |
+
logger.info("Upstream streaming response closed.")
|
163 |
+
|
164 |
+
def format_messages_for_xplain(messages) -> str:
|
165 |
+
"""格式化消息列表为DeepSider API所需的提示格式"""
|
166 |
+
prompt = ""
|
167 |
+
for msg in messages:
|
168 |
+
role = msg["role"]
|
169 |
+
# 将OpenAI的角色映射到DeepSider能理解的格式
|
170 |
+
if role == "system":
|
171 |
+
# 系统消息放在开头 作为指导
|
172 |
+
prompt = f"{msg['content']}\n\n" + prompt
|
173 |
+
elif role == "user":
|
174 |
+
prompt += f"Human: {msg['content']}\n\n"
|
175 |
+
elif role == "assistant":
|
176 |
+
content = msg['content']
|
177 |
+
prompt += f"Assistant: {msg['content']}\n\n"
|
178 |
+
|
179 |
+
else:
|
180 |
+
# 其他角色按用户处理
|
181 |
+
prompt += f"Human ({role}): {msg['content']}\n\n"
|
182 |
+
|
183 |
+
# 如果最后一个消息不是用户的 添加一个Human前缀引导模型回答
|
184 |
+
if messages and messages[-1]["role"] != "user":
|
185 |
+
prompt += "Human: "
|
186 |
+
|
187 |
+
return prompt.strip()
|
188 |
+
|
189 |
+
|
190 |
+
# --- Proxy Endpoint ---
|
191 |
+
@app.post("/v1/chat/completions")
|
192 |
+
async def proxy_openai_chat(request: Request, _client_key: str = Depends(verify_api_key)): # Use Depends for auth
|
193 |
+
"""
|
194 |
+
Proxies requests to the configured Chat Completions endpoint AFTER verifying client API key.
|
195 |
+
Uses rotated keys for upstream authentication.
|
196 |
+
"""
|
197 |
+
client: httpx.AsyncClient = request.app.state.http_client
|
198 |
+
key_iterator = request.app.state.key_iterator
|
199 |
+
|
200 |
+
if not client or not key_iterator:
|
201 |
+
logger.error("HTTPX client or Key Iterator not available (app state issue).")
|
202 |
+
raise HTTPException(status_code=503, detail="Service temporarily unavailable")
|
203 |
+
|
204 |
+
# --- Get Next Upstream API Key ---
|
205 |
+
try:
|
206 |
+
current_upstream_key = next(key_iterator)
|
207 |
+
# Log rotation (optional, consider security of logging key info)
|
208 |
+
# logger.info(f"Using upstream key ending: ...{current_upstream_key[-4:]}")
|
209 |
+
except StopIteration:
|
210 |
+
# This should not happen if lifespan logic is correct and keys were loaded
|
211 |
+
logger.error("Upstream key iterator exhausted unexpectedly.")
|
212 |
+
raise HTTPException(status_code=500, detail="Internal Server Error: Key rotation failed")
|
213 |
+
except Exception as e:
|
214 |
+
logger.error(f"Unexpected error getting next key: {e}")
|
215 |
+
raise HTTPException(status_code=500, detail="Internal Server Error: Key rotation failed")
|
216 |
+
|
217 |
+
# --- Get Request Data ---
|
218 |
+
try:
|
219 |
+
request_body = await request.body()
|
220 |
+
payload = json.loads(request_body)
|
221 |
+
except json.JSONDecodeError:
|
222 |
+
raise HTTPException(status_code=400, detail="Invalid JSON body")
|
223 |
+
type = payload.get("type", "claude")
|
224 |
+
|
225 |
+
is_streaming = payload.get("stream", False)
|
226 |
+
|
227 |
+
# --- Prepare Upstream Request ---
|
228 |
+
upstream_headers = {
|
229 |
+
"Content-Type": request.headers.get("Content-Type", "application/json"),
|
230 |
+
"Accept": request.headers.get("Accept", "application/json"),
|
231 |
+
}
|
232 |
+
|
233 |
+
# --- Upstream Authentication (Using Rotated Key) ---
|
234 |
+
# Decide based on the target API (e.g., freeaichatplayground vs standard OpenAI)
|
235 |
+
if "freeaichatplayground.com" in OPENAI_API_BASE:
|
236 |
+
logger.debug("Using payload apiKey for upstream authentication (freeaichatplayground specific).")
|
237 |
+
payload["apiKey"] = current_upstream_key # Inject ROTATED key into payload
|
238 |
+
else:
|
239 |
+
# Default to standard Bearer token authentication for upstream
|
240 |
+
logger.debug("Using Authorization header for upstream authentication.")
|
241 |
+
upstream_headers["Authorization"] = f"Bearer {current_upstream_key}" # Use ROTATED key
|
242 |
+
upstream_headers["x-api-key"] = f"{current_upstream_key}" # Use ROTATED key
|
243 |
+
|
244 |
+
if is_streaming and "text/event-stream" not in upstream_headers["Accept"]:
|
245 |
+
logger.info("Adding 'Accept: text/event-stream' for streaming request")
|
246 |
+
upstream_headers["Accept"] = "text/event-stream, application/json"
|
247 |
+
|
248 |
+
logger.info(f"Forwarding request to {OPENAI_CHAT_ENDPOINT} (Streaming: {is_streaming})")
|
249 |
+
|
250 |
+
|
251 |
+
|
252 |
+
# --- Make Request to Upstream ---
|
253 |
+
response = None # Define response here to ensure it's available in finally block
|
254 |
+
response_body = None
|
255 |
+
try:
|
256 |
+
if type == "claude":
|
257 |
+
newPayload = {
|
258 |
+
"text": format_messages_for_xplain(payload["messages"]),
|
259 |
+
}
|
260 |
+
req = client.build_request("POST","https://models.aixplain.com/api/v1/execute/" + payload["model"],json=newPayload,headers=upstream_headers)
|
261 |
+
response = await client.send(req)
|
262 |
+
xreqId = response.json()["data"]
|
263 |
+
|
264 |
+
req = client.build_request("GET",xreqId)
|
265 |
+
tryCount = 0
|
266 |
+
while tryCount < 100:
|
267 |
+
response = await client.send(req)
|
268 |
+
tryCount += 1
|
269 |
+
|
270 |
+
rj = response.json()
|
271 |
+
if rj["completed"] == True:
|
272 |
+
data = {"id": "123456-456789-123456", "object": "chat.completion.chunk",
|
273 |
+
"choices": [{"delta": {"content": rj["data"]}, "index": 0, "finish_reason": None}]}
|
274 |
+
response_body = json.dumps(data)
|
275 |
+
break
|
276 |
+
time.sleep(1)
|
277 |
+
else:
|
278 |
+
req = client.build_request(
|
279 |
+
"POST",
|
280 |
+
OPENAI_CHAT_ENDPOINT, # Use the full URL
|
281 |
+
json=payload,
|
282 |
+
headers=upstream_headers,
|
283 |
+
)
|
284 |
+
response = await client.send(req, stream=True)
|
285 |
+
if response.status_code >= 400:
|
286 |
+
error_body = await response.aread() # Read error fully
|
287 |
+
await response.aclose()
|
288 |
+
logger.error(f"Upstream API returned error: {response.status_code} Key ending: ...{current_upstream_key[-4:]} Body: {error_body.decode()}")
|
289 |
+
try: detail = json.loads(error_body)
|
290 |
+
except json.JSONDecodeError: detail = error_body.decode()
|
291 |
+
raise HTTPException(status_code=response.status_code, detail=detail)
|
292 |
+
response_body = await response.aread()
|
293 |
+
response_body = response_body.decode()
|
294 |
+
|
295 |
+
# --- Handle Streaming Response ---
|
296 |
+
if is_streaming:
|
297 |
+
logger.info(f"Received OK streaming response from upstream (Status: {response.status_code}). Piping to client.")
|
298 |
+
status_code = response.status_code
|
299 |
+
if status_code == 201:
|
300 |
+
status_code = 200
|
301 |
+
return StreamingResponse(
|
302 |
+
yield_openai_chunks(response_body), # Generator handles closing response
|
303 |
+
status_code=status_code,
|
304 |
+
media_type=response.headers.get("content-type", "text/event-stream"),
|
305 |
+
)
|
306 |
+
# --- Handle Non-Streaming Response ---
|
307 |
+
else:
|
308 |
+
logger.info(f"Received OK non-streaming response from upstream (Status: {response.status_code}). Reading full body.")
|
309 |
+
response_body = await response.aread()
|
310 |
+
await response.aclose() # Ensure closed
|
311 |
+
content_type = response.headers.get("content-type", "application/json")
|
312 |
+
return Response( # Return raw response, FastAPI handles JSON content type
|
313 |
+
content=response_body,
|
314 |
+
status_code=response.status_code,
|
315 |
+
media_type=content_type,
|
316 |
+
)
|
317 |
+
|
318 |
+
except httpx.TimeoutException as e:
|
319 |
+
logger.error(f"Request to upstream timed out: {e}")
|
320 |
+
if response: await response.aclose()
|
321 |
+
raise HTTPException(status_code=504, detail="Request to upstream API timed out.")
|
322 |
+
except httpx.RequestError as e:
|
323 |
+
logger.error(f"Error requesting upstream API: {e}")
|
324 |
+
if response: await response.aclose()
|
325 |
+
raise HTTPException(status_code=502, detail=f"Error contacting upstream API: {e}")
|
326 |
+
except HTTPException as e:
|
327 |
+
# Re-raise FastAPI HTTPExceptions (like the 4xx check above)
|
328 |
+
if response and not response.is_closed: await response.aclose()
|
329 |
+
raise e
|
330 |
+
except Exception as e:
|
331 |
+
logger.exception("An unexpected error occurred during response processing.")
|
332 |
+
if response and not response.is_closed: await response.aclose()
|
333 |
+
raise HTTPException(status_code=500, detail=f"Internal server error: {str(e)}")
|
334 |
+
|
335 |
+
|
336 |
+
# --- Health Check Endpoint ---
|
337 |
+
@app.get("/health")
|
338 |
+
async def health_check():
|
339 |
+
"""Simple health check endpoint."""
|
340 |
+
# Could add checks here, e.g., if keys were loaded
|
341 |
+
key_count = len(app.state.upstream_api_keys) if hasattr(app.state, 'upstream_api_keys') else 0
|
342 |
+
return {"status": "ok", "upstream_keys_loaded": key_count > 0, "key_count": key_count}
|
343 |
+
|
344 |
+
# --- Main Execution Guard ---
|
345 |
+
if __name__ == "__main__":
|
346 |
+
import uvicorn
|
347 |
+
# Startup checks are implicitly handled by config loading at the top
|
348 |
+
print("--- Starting FastAPI OpenAI Proxy with Custom Auth & Key Rotation ---")
|
349 |
+
print(f"Proxying requests to: {OPENAI_CHAT_ENDPOINT}")
|
350 |
+
if UPSTREAM_KEYS_URL:
|
351 |
+
print(f"Fetching upstream keys from: {UPSTREAM_KEYS_URL}")
|
352 |
+
elif DEFAULT_OPENAI_API_KEY:
|
353 |
+
print("Using single OPENAI_API_KEY for upstream.")
|
354 |
+
else:
|
355 |
+
print("ERROR: No upstream key source configured!") # Should have failed earlier
|
356 |
+
print(f"Clients must provide a valid API key in 'Authorization: Bearer <key>' header.")
|
357 |
+
print(f"Number of allowed client keys configured: {len(ALLOWED_KEYS)}")
|
358 |
+
print("---")
|
359 |
+
|
360 |
+
uvicorn.run(app, host="0.0.0.0", port=7860)
|
361 |
+
|