Update api/utils.py

api/utils.py

@@ -9,49 +9,25 @@ from typing import Any, Dict, Optional
 import httpx
 from fastapi import HTTPException
 from api.config import (
-    MODEL_MAPPING,
+    AGENT_MODE,
+    TRENDING_AGENT_MODE,
     get_headers_api_chat,
     get_headers_chat,
     BASE_URL,
-    AGENT_MODE,
-    TRENDING_AGENT_MODE,
-    MODEL_PREFIXES,
-    MODEL_REFERERS
 )
 from api.models import ChatRequest
 from api.logger import setup_logger
 
 logger = setup_logger(__name__)
 
-# Helper function to create a random alphanumeric chat ID
+# Updated message ID generator
 def generate_chat_id(length: int = 7) -> str:
     characters = string.ascii_letters + string.digits
-    return ''.join(random.choices(characters, k=length))
-
-# Helper function to create chat completion data
-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,
-    }
+    return ''.join(random.choice(characters) for _ in range(length))
 
-# Function to convert message to dictionary format, ensuring base64 data and optional model prefix
-def message_to_dict(message, model_prefix: Optional[str] = None):
+# Updated message_to_dict function (removed model prefixes)
+def message_to_dict(message):
     content = message.content if isinstance(message.content, str) else message.content[0]["text"]
-    if model_prefix:
-        content = f"{model_prefix} {content}"
     if isinstance(message.content, list) and len(message.content) == 2 and "image_url" in message.content[1]:
         # Ensure base64 images are always included for all models
         return {
@@ -65,22 +41,16 @@ def message_to_dict(message, model_prefix: Optional[str] = None):
         }
     return {"role": message.role, "content": content}
 
-# Function to strip model prefix from content if present
-def strip_model_prefix(content: str, model_prefix: Optional[str] = None) -> str:
-    """Remove the model prefix from the response content if present."""
-    if model_prefix and content.startswith(model_prefix):
-        logger.debug(f"Stripping prefix '{model_prefix}' from content.")
-        return content[len(model_prefix):].strip()
-    return content
+# Removed strip_model_prefix function as per your requirement
 
-# Function to get the correct referer URL for logging
+# Function to get the correct referer URL
 def get_referer_url(chat_id: str, model: str) -> str:
     """Generate the referer URL based on specific models listed in MODEL_REFERERS."""
     if model in MODEL_REFERERS:
-        return f"{BASE_URL}/chat/{chat_id}?model={model}"
+        return f"{BASE_URL}{MODEL_REFERERS[model]}"
     return BASE_URL
 
-# Process streaming response with headers from config.py
+# Process streaming response with updated headers and data
 async def process_streaming_response(request: ChatRequest):
     chat_id = generate_chat_id()
     referer_url = get_referer_url(chat_id, request.model)
@@ -88,37 +58,32 @@ async def process_streaming_response(request: ChatRequest):
 
     agent_mode = AGENT_MODE.get(request.model, {})
     trending_agent_mode = TRENDING_AGENT_MODE.get(request.model, {})
-    model_prefix = MODEL_PREFIXES.get(request.model, "")
 
     headers_api_chat = get_headers_api_chat(referer_url)
 
-    if request.model == 'o1-preview':
-        delay_seconds = random.randint(1, 60)
-        logger.info(f"Introducing a delay of {delay_seconds} seconds for model 'o1-preview' (Chat ID: {chat_id})")
-        await asyncio.sleep(delay_seconds)
-
     json_data = {
+        "messages": [message_to_dict(msg) for msg in request.messages],
+        "id": chat_id,
+        "previewToken": None,
+        "userId": None,
+        "codeModelMode": True,
         "agentMode": agent_mode,
+        "trendingAgentMode": trending_agent_mode,
+        "isMicMode": False,
+        "userSystemPrompt": None,
+        "maxTokens": request.max_tokens,
+        "playgroundTopP": request.top_p,
+        "playgroundTemperature": request.temperature,
+        "isChromeExt": False,
+        "githubToken": None,
         "clickedAnswer2": False,
         "clickedAnswer3": False,
         "clickedForceWebSearch": False,
-        "codeModelMode": True,
-        "githubToken": None,
-        "id": chat_id,
-        "isChromeExt": False,
-        "isMicMode": False,
-        "maxTokens": request.max_tokens,
-        "messages": [message_to_dict(msg, model_prefix=model_prefix) for msg in request.messages],
+        "visitFromDelta": False,
         "mobileClient": False,
-        "playgroundTemperature": request.temperature,
-        "playgroundTopP": request.top_p,
-        "previewToken": None,
-        "trendingAgentMode": trending_agent_mode,
-        "userId": None,
-        "userSelectedModel": MODEL_MAPPING.get(request.model, request.model),
-        "userSystemPrompt": None,
+        "userSelectedModel": request.model if request.model in ["gpt-4o", "gemini-pro", "claude-sonnet-3.5", "blackboxai-pro"] else None,
+        "webSearchMode": False,  # Adjust if needed
         "validated": "69783381-2ce4-4dbd-ac78-35e9063feabc",
-        "visitFromDelta": False,
     }
 
     async with httpx.AsyncClient() as client:
@@ -134,11 +99,9 @@ async def process_streaming_response(request: ChatRequest):
                 async for line in response.aiter_lines():
                     timestamp = int(datetime.now().timestamp())
                     if line:
-                        content = line
-                        if content.startswith("$@$v=undefined-rv1$@$"):
-                            content = content[21:]
-                        cleaned_content = strip_model_prefix(content, model_prefix)
-                        yield f"data: {json.dumps(create_chat_completion_data(cleaned_content, request.model, timestamp))}\n\n"
+                        content = line.strip()
+                        # Handle special cases if necessary
+                        yield f"data: {json.dumps(create_chat_completion_data(content, request.model, timestamp))}\n\n"
 
                 yield f"data: {json.dumps(create_chat_completion_data('', request.model, timestamp, 'stop'))}\n\n"
                 yield "data: [DONE]\n\n"
@@ -149,7 +112,7 @@ async def process_streaming_response(request: ChatRequest):
             logger.error(f"Error occurred during request for Chat ID {chat_id}: {e}")
             raise HTTPException(status_code=500, detail=str(e))
 
-# Process non-streaming response with headers from config.py
+# Process non-streaming response with updated headers and data
 async def process_non_streaming_response(request: ChatRequest):
     chat_id = generate_chat_id()
     referer_url = get_referer_url(chat_id, request.model)
@@ -157,38 +120,33 @@ async def process_non_streaming_response(request: ChatRequest):
 
     agent_mode = AGENT_MODE.get(request.model, {})
     trending_agent_mode = TRENDING_AGENT_MODE.get(request.model, {})
-    model_prefix = MODEL_PREFIXES.get(request.model, "")
 
     headers_api_chat = get_headers_api_chat(referer_url)
     headers_chat = get_headers_chat(referer_url, next_action=str(uuid.uuid4()), next_router_state_tree=json.dumps([""]))
 
-    if request.model == 'o1-preview':
-        delay_seconds = random.randint(20, 60)
-        logger.info(f"Introducing a delay of {delay_seconds} seconds for model 'o1-preview' (Chat ID: {chat_id})")
-        await asyncio.sleep(delay_seconds)
-
     json_data = {
+        "messages": [message_to_dict(msg) for msg in request.messages],
+        "id": chat_id,
+        "previewToken": None,
+        "userId": None,
+        "codeModelMode": True,
         "agentMode": agent_mode,
+        "trendingAgentMode": trending_agent_mode,
+        "isMicMode": False,
+        "userSystemPrompt": None,
+        "maxTokens": request.max_tokens,
+        "playgroundTopP": request.top_p,
+        "playgroundTemperature": request.temperature,
+        "isChromeExt": False,
+        "githubToken": None,
         "clickedAnswer2": False,
         "clickedAnswer3": False,
         "clickedForceWebSearch": False,
-        "codeModelMode": True,
-        "githubToken": None,
-        "id": chat_id,
-        "isChromeExt": False,
-        "isMicMode": False,
-        "maxTokens": request.max_tokens,
-        "messages": [message_to_dict(msg, model_prefix=model_prefix) for msg in request.messages],
+        "visitFromDelta": False,
         "mobileClient": False,
-        "playgroundTemperature": request.temperature,
-        "playgroundTopP": request.top_p,
-        "previewToken": None,
-        "trendingAgentMode": trending_agent_mode,
-        "userId": None,
-        "userSelectedModel": MODEL_MAPPING.get(request.model, request.model),
-        "userSystemPrompt": None,
+        "userSelectedModel": request.model if request.model in ["gpt-4o", "gemini-pro", "claude-sonnet-3.5", "blackboxai-pro"] else None,
+        "webSearchMode": False,  # Adjust if needed
         "validated": "69783381-2ce4-4dbd-ac78-35e9063feabc",
-        "visitFromDelta": False,
     }
 
     full_response = ""
@@ -199,17 +157,13 @@ async def process_non_streaming_response(request: ChatRequest):
             ) as response:
                 response.raise_for_status()
                 async for chunk in response.aiter_text():
-                    full_response += chunk
+                    full_response += chunk.strip()
         except httpx.HTTPStatusError as e:
             logger.error(f"HTTP error occurred for Chat ID {chat_id}: {e}")
             raise HTTPException(status_code=e.response.status_code, detail=str(e))
         except httpx.RequestError as e:
             logger.error(f"Error occurred during request for Chat ID {chat_id}: {e}")
             raise HTTPException(status_code=500, detail=str(e))
-    if full_response.startswith("$@$v=undefined-rv1$@$"):
-        full_response = full_response[21:]
-
-    cleaned_full_response = strip_model_prefix(full_response, model_prefix)
 
     return {
         "id": f"chatcmpl-{uuid.uuid4()}",
@@ -219,7 +173,7 @@ async def process_non_streaming_response(request: ChatRequest):
         "choices": [
             {
                 "index": 0,
-                "message": {"role": "assistant", "content": cleaned_full_response},
+                "message": {"role": "assistant", "content": full_response},
                 "finish_reason": "stop",
             }
         ],