File size: 8,951 Bytes
a4b70d9 |
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 |
from __future__ import annotations
import json
import base64
import time
import random
import hashlib
import asyncio
from aiohttp import ClientSession
from ...typing import AsyncResult, Messages
from ...errors import ResponseError
from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin
from ..helper import format_prompt, get_last_user_message
from ...providers.response import FinishReason, JsonConversation
class Conversation(JsonConversation):
message_history: Messages = []
def __init__(self, model: str):
self.model = model
self.message_history = []
class DDG(AsyncGeneratorProvider, ProviderModelMixin):
label = "DuckDuckGo AI Chat"
url = "https://duckduckgo.com"
api_endpoint = "https://duckduckgo.com/duckchat/v1/chat"
status_url = "https://duckduckgo.com/duckchat/v1/status"
working = False
needs_auth = False
supports_stream = True
supports_system_message = True
supports_message_history = True
default_model = "gpt-4o-mini"
model_aliases = {
"gpt-4": default_model,
"gpt-4o": default_model,
"llama-3.3-70b": "meta-llama/Llama-3.3-70B-Instruct-Turbo",
"claude-3-haiku": "claude-3-haiku-20240307",
"mistral-small": "mistralai/Mistral-Small-24B-Instruct-2501",
"mistral-small-24b": "mistralai/Mistral-Small-24B-Instruct-2501",
}
models = [default_model, "o3-mini"] + list(model_aliases.keys())
@staticmethod
def generate_user_agent() -> str:
return f"Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.{random.randint(1000,9999)}.0 Safari/537.36"
@staticmethod
def generate_fe_signals() -> str:
current_time = int(time.time() * 1000)
signals_data = {
"start": current_time - 35000,
"events": [
{"name": "onboarding_impression_1", "delta": 383},
{"name": "onboarding_impression_2", "delta": 6004},
{"name": "onboarding_finish", "delta": 9690},
{"name": "startNewChat", "delta": 10082},
{"name": "initSwitchModel", "delta": 16586}
],
"end": 35163
}
return base64.b64encode(json.dumps(signals_data).encode()).decode()
@staticmethod
def generate_fe_version(page_content: str = "") -> str:
try:
fe_hash = page_content.split('__DDG_FE_CHAT_HASH__="', 1)[1].split('"', 1)[0]
return f"serp_20250510_052906_ET-{fe_hash}"
except Exception:
return "serp_20250510_052906_ET-ed4f51dc2e106020bc4b"
@staticmethod
def generate_x_vqd_hash_1(vqd: str, fe_version: str) -> str:
# Placeholder logic; in reality DuckDuckGo uses dynamic JS challenge
concat = f"{vqd}#{fe_version}"
hash_digest = hashlib.sha256(concat.encode()).digest()
b64 = base64.b64encode(hash_digest).decode()
return base64.b64encode(json.dumps({
"server_hashes": [],
"client_hashes": [b64],
"signals": {},
"meta": {
"v": "1",
"challenge_id": hashlib.md5(concat.encode()).hexdigest(),
"origin": "https://duckduckgo.com",
"stack": "Generated in Python"
}
}).encode()).decode()
@classmethod
async def create_async_generator(
cls,
model: str,
messages: Messages,
proxy: str = None,
conversation: Conversation = None,
return_conversation: bool = True,
retry_count: int = 0,
**kwargs
) -> AsyncResult:
model = cls.get_model(model)
if conversation is None:
conversation = Conversation(model)
conversation.message_history = messages.copy()
else:
last_message = next((m for m in reversed(messages) if m["role"] == "user"), None)
if last_message and last_message not in conversation.message_history:
conversation.message_history.append(last_message)
base_headers = {
"accept-language": "en-US,en;q=0.9",
"dnt": "1",
"origin": "https://duckduckgo.com",
"referer": "https://duckduckgo.com/",
"sec-ch-ua": '"Chromium";v="135", "Not-A.Brand";v="8"',
"sec-ch-ua-mobile": "?0",
"sec-ch-ua-platform": '"Linux"',
"sec-fetch-dest": "empty",
"sec-fetch-mode": "cors",
"sec-fetch-site": "same-origin",
"user-agent": cls.generate_user_agent(),
}
cookies = {'dcs': '1', 'dcm': '3'}
formatted_prompt = format_prompt(conversation.message_history) if len(conversation.message_history) > 1 else get_last_user_message(messages)
data = {"model": model, "messages": [{"role": "user", "content": formatted_prompt}], "canUseTools": False}
async with ClientSession(cookies=cookies) as session:
try:
# Step 1: Initial page load
async with session.get(f"{cls.url}/?q=DuckDuckGo+AI+Chat&ia=chat&duckai=1",
headers={**base_headers, "accept": "text/html"}, proxy=proxy) as r:
r.raise_for_status()
page = await r.text()
fe_version = cls.generate_fe_version(page)
# Step 2: Get VQD
status_headers = {**base_headers, "accept": "*/*", "cache-control": "no-store", "x-vqd-accept": "1"}
async with session.get(cls.status_url, headers=status_headers, proxy=proxy) as r:
r.raise_for_status()
vqd = r.headers.get("x-vqd-4", "") or f"4-{random.randint(10**29, 10**30 - 1)}"
x_vqd_hash_1 = cls.generate_x_vqd_hash_1(vqd, fe_version)
# Step 3: Actual chat request
chat_headers = {
**base_headers,
"accept": "text/event-stream",
"content-type": "application/json",
"x-fe-signals": cls.generate_fe_signals(),
"x-fe-version": fe_version,
"x-vqd-4": vqd,
"x-vqd-hash-1": x_vqd_hash_1,
}
async with session.post(cls.api_endpoint, json=data, headers=chat_headers, proxy=proxy) as response:
if response.status != 200:
error_text = await response.text()
if "ERR_BN_LIMIT" in error_text:
raise ResponseError("Blocked by DuckDuckGo: Bot limit exceeded (ERR_BN_LIMIT).")
if "ERR_INVALID_VQD" in error_text and retry_count < 3:
await asyncio.sleep(random.uniform(2.5, 5.5))
async for chunk in cls.create_async_generator(
model, messages, proxy, conversation, return_conversation, retry_count + 1, **kwargs
):
yield chunk
return
raise ResponseError(f"HTTP {response.status} - {error_text}")
full_message = ""
async for line in response.content:
line_text = line.decode("utf-8").strip()
if line_text.startswith("data:"):
payload = line_text[5:].strip()
if payload == "[DONE]":
if full_message:
conversation.message_history.append({"role": "assistant", "content": full_message})
if return_conversation:
yield conversation
yield FinishReason("stop")
break
try:
msg = json.loads(payload)
if msg.get("action") == "error":
raise ResponseError(f"Error: {msg.get('type', 'unknown')}")
if "message" in msg:
content = msg["message"]
yield content
full_message += content
except json.JSONDecodeError:
continue
except Exception as e:
if retry_count < 3:
await asyncio.sleep(random.uniform(2.5, 5.5))
async for chunk in cls.create_async_generator(
model, messages, proxy, conversation, return_conversation, retry_count + 1, **kwargs
):
yield chunk
else:
raise ResponseError(f"Error: {str(e)}")
|