File size: 9,274 Bytes
14cd2f6
 
0125000
9edc2f1
14cd2f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
08606b0
5bf1b3e
 
14cd2f6
5bf1b3e
 
9edc2f1
14cd2f6
9edc2f1
5bf1b3e
 
14cd2f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
745d3f1
14cd2f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9edc2f1
14cd2f6
 
745d3f1
14cd2f6
 
 
 
9edc2f1
14cd2f6
 
 
 
9edc2f1
14cd2f6
 
 
 
9edc2f1
14cd2f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9edc2f1
14cd2f6
 
 
 
9edc2f1
14cd2f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
745d3f1
 
 
14cd2f6
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
from fastapi import FastAPI, Request, HTTPException
from fastapi.responses import JSONResponse, StreamingResponse
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from typing import List, Dict, Any, Union
import os
import time
import httpx
import json
from dotenv import load_dotenv

load_dotenv()

# Simple configuration
API_PREFIX = os.getenv("API_PREFIX", "/")
MAX_RETRY_COUNT = int(os.getenv("MAX_RETRY_COUNT", "3"))
RETRY_DELAY = int(os.getenv("RETRY_DELAY", "5000"))

# Default headers for DuckDuckGo requests
FAKE_HEADERS = {
    "Accept": "*/*",
    "Accept-Language": "en-US,en;q=0.9",
    "Origin": "https://duckduckgo.com/",
    "Referer": "https://duckduckgo.com/",
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36",
}

app = FastAPI()

# Add CORS middleware
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_methods=["*"],
    allow_headers=["*"],
)

# Models for request validation
class Message(BaseModel):
    role: str
    content: Union[str, List[Dict[str, Any]]]

class ChatCompletionRequest(BaseModel):
    model: str
    messages: List[Message]
    stream: bool = False

# Add timing information
@app.middleware("http")
async def add_process_time(request: Request, call_next):
    start_time = time.time()
    response = await call_next(request)
    process_time = time.time() - start_time
    print(f"{request.method} {response.status_code} {request.url.path} {process_time*1000:.2f} ms")
    return response

@app.get("/")
async def root():
    return {"message": "API server running"}

@app.get("/ping")
async def ping():
    return {"message": "pong"}

@app.get(f"{API_PREFIX}v1/models")
async def get_models():
    return {
        "object": "list",
        "data": [
            {"id": "gpt-4o-mini", "object": "model", "owned_by": "ddg"},
            {"id": "claude-3-haiku", "object": "model", "owned_by": "ddg"},
            {"id": "llama-3.1-70b", "object": "model", "owned_by": "ddg"},
            {"id": "mixtral-8x7b", "object": "model", "owned_by": "ddg"},
            {"id": "o3-mini", "object": "model", "owned_by": "ddg"},
        ],
    }

@app.post(f"{API_PREFIX}v1/chat/completions")
async def chat_completions(request: ChatCompletionRequest):
    try:
        model = convert_model(request.model)
        content = messages_to_text(request.messages)
        return await create_completion(model, content, request.stream)
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

def convert_model(input_model: str) -> str:
    """Convert public model names to DuckDuckGo internal model names"""
    model_mapping = {
        "claude-3-haiku": "claude-3-haiku-20240307",
        "llama-3.1-70b": "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
        "mixtral-8x7b": "mistralai/Mixtral-8x7B-Instruct-v0.1",
        "o3-mini": "o3-mini"
    }
    return model_mapping.get(input_model.lower(), "gpt-4o-mini")

def messages_to_text(messages: List[Message]) -> str:
    """Convert message array to text format expected by DuckDuckGo API"""
    result = ""
    for message in messages:
        role = "user" if message.role == "system" else message.role
        
        if role in ["user", "assistant"]:
            # Handle both string content and structured content
            if isinstance(message.content, list):
                content_str = "".join([item.get("text", "") for item in message.content if item.get("text", "")])
            else:
                content_str = message.content
            
            result += f"{role}:{content_str};\r\n"
    
    return result

async def request_token() -> str:
    """Get auth token from DuckDuckGo"""
    try:
        async with httpx.AsyncClient() as client:
            response = await client.get(
                "https://duckduckgo.com/duckchat/v1/status",
                headers={**FAKE_HEADERS, "x-vqd-accept": "1"}
            )
            return response.headers.get("x-vqd-4", "")
    except Exception as e:
        print(f"Token request error: {e}")
        return ""

async def create_completion(model: str, content: str, return_stream: bool, retry_count: int = 0):
    """Create a chat completion via DuckDuckGo API"""
    token = await request_token()
    
    try:
        async with httpx.AsyncClient() as client:
            response = await client.post(
                "https://duckduckgo.com/duckchat/v1/chat",
                headers={
                    **FAKE_HEADERS,
                    "Accept": "text/event-stream",
                    "Content-Type": "application/json",
                    "x-vqd-4": token,
                },
                json={
                    "model": model,
                    "messages": [{"role": "user", "content": content}]
                },
                stream=True
            )

            if response.status_code != 200:
                raise HTTPException(status_code=response.status_code, detail="API request failed")
            
            return await process_stream(model, response, return_stream)
    except Exception as e:
        if retry_count < MAX_RETRY_COUNT:
            print(f"Retrying... attempt {retry_count + 1}")
            await asyncio.sleep(RETRY_DELAY / 1000)
            return await create_completion(model, content, return_stream, retry_count + 1)
        raise HTTPException(status_code=500, detail=str(e))

async def process_stream(model: str, response, return_stream: bool):
    """Process streaming response from DuckDuckGo"""
    buffer = ""
    full_text = ""
    
    async def generate_stream():
        nonlocal buffer, full_text
        
        # Process chunks as they arrive
        async for chunk in response.aiter_bytes():
            chunk_str = chunk.decode('utf-8').strip()
            
            # Handle buffer from previous chunk if needed
            if buffer:
                chunk_str = buffer + chunk_str
                buffer = ""
            
            # Handle incomplete chunks
            if not chunk_str.endswith('"}') and "[DONE]" not in chunk_str:
                buffer = chunk_str
                continue
            
            # Process each line in the chunk
            for line in chunk_str.split('\n'):
                if len(line) < 6:
                    continue
                
                # Remove prefix (data: )
                line = line[6:] if line.startswith("data: ") else line
                
                # Handle completion signal
                if line == "[DONE]":
                    if return_stream:
                        yield f"data: {json.dumps(create_stop_chunk(model))}\n\n"
                    return
                
                # Parse and handle message content
                try:
                    data = json.loads(line)
                    if data.get("action") == "success" and "message" in data:
                        message = data["message"]
                        full_text += message
                        
                        if return_stream:
                            yield f"data: {json.dumps(create_chunk(message, model))}\n\n"
                except json.JSONDecodeError:
                    continue

    # Return appropriate response based on streaming preference
    if return_stream:
        return StreamingResponse(generate_stream(), media_type="text/event-stream")
    else:
        # For non-streaming, consume the generator and return complete response
        async for _ in generate_stream():
            pass  # Just collecting text in full_text
        
        return JSONResponse(content=create_complete_response(full_text, model))

def create_chunk(text: str, model: str) -> dict:
    """Create a streaming chunk response"""
    return {
        "id": "chatcmpl-123",
        "object": "chat.completion.chunk",
        "created": int(time.time()),
        "model": model,
        "choices": [
            {
                "index": 0,
                "delta": {"content": text},
                "finish_reason": None,
            },
        ],
    }

def create_stop_chunk(model: str) -> dict:
    """Create a final streaming chunk with stop reason"""
    return {
        "id": "chatcmpl-123",
        "object": "chat.completion.chunk",
        "created": int(time.time()),
        "model": model,
        "choices": [
            {
                "index": 0,
                "delta": {},
                "finish_reason": "stop",
            },
        ],
    }

def create_complete_response(text: str, model: str) -> dict:
    """Create a complete non-streaming response"""
    return {
        "id": "chatcmpl-123",
        "object": "chat.completion",
        "created": int(time.time()),
        "model": model,
        "usage": {"prompt_tokens": 0, "completion_tokens": 0, "total_tokens": 0},
        "choices": [
            {
                "message": {"content": text, "role": "assistant"},
                "index": 0,
                "finish_reason": "stop",
            },
        ],
    }

# Only needed for retry delays
import asyncio

if __name__ == "__main__":
    import uvicorn
    uvicorn.run("app:app", host="0.0.0.0", port=7860, reload=True)