Add full logging

main.py

@@ -17,6 +17,18 @@ from prompts import CODING_ASSISTANT_PROMPT, NEWS_ASSISTANT_PROMPT, generate_new
 from fastapi_cache import FastAPICache
 from fastapi_cache.backends.inmemory import InMemoryBackend
 from fastapi_cache.decorator import cache
+import logging
+
+# Configure logging
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    handlers=[
+        logging.FileHandler("app.log"),
+        logging.StreamHandler()
+    ]
+)
+logger = logging.getLogger(__name__)
 
 app = FastAPI()
 
@@ -69,6 +81,7 @@ class NewsQueryModel(BaseModel):
 
 @lru_cache()
 def get_api_keys():
+    logger.info("Loading API keys")
     return {
         "OPENROUTER_API_KEY": f"sk-or-v1-{os.environ['OPENROUTER_API_KEY']}",
         "BRAVE_API_KEY": os.environ['BRAVE_API_KEY']
@@ -91,6 +104,7 @@ def calculate_tokens(msgs):
     return sum(len(encoding.encode(str(m))) for m in msgs)
 
 def chat_with_llama_stream(messages, model="gpt-3.5-turbo", max_llm_history=4, max_output_tokens=2500):
+    logger.info(f"Starting chat with model: {model}")
     while calculate_tokens(messages) > (8000 - max_output_tokens):
         if len(messages) > max_llm_history:
             messages = [messages[0]] + messages[-max_llm_history:]
@@ -98,6 +112,7 @@ def chat_with_llama_stream(messages, model="gpt-3.5-turbo", max_llm_history=4, m
             max_llm_history -= 1
             if max_llm_history < 2:
                 error_message = "Token limit exceeded. Please shorten your input or start a new conversation."
+                logger.error(error_message)
                 raise HTTPException(status_code=400, detail=error_message)
 
     try:
@@ -117,11 +132,14 @@ def chat_with_llama_stream(messages, model="gpt-3.5-turbo", max_llm_history=4, m
         
         # After streaming, add the full response to the conversation history
         messages.append({"role": "assistant", "content": full_response})
+        logger.info("Chat completed successfully")
     except Exception as e:
+        logger.error(f"Error in model response: {str(e)}")
         raise HTTPException(status_code=500, detail=f"Error in model response: {str(e)}")
 
 async def verify_api_key(api_key: str = Security(api_key_header)):
     if api_key != API_KEY:
+        logger.warning("Invalid API key used")
         raise HTTPException(status_code=403, detail="Could not validate credentials")
     return api_key
 
@@ -129,6 +147,7 @@ async def verify_api_key(api_key: str = Security(api_key_header)):
 DB_PATH = '/app/data/conversations.db'
 
 def init_db():
+    logger.info("Initializing database")
     os.makedirs(os.path.dirname(DB_PATH), exist_ok=True)
     conn = sqlite3.connect(DB_PATH)
     c = conn.cursor()
@@ -141,19 +160,23 @@ def init_db():
                   timestamp DATETIME DEFAULT CURRENT_TIMESTAMP)''')
     conn.commit()
     conn.close()
+    logger.info("Database initialized successfully")
 
 init_db()
 
 def update_db(user_id, conversation_id, message, response):
+    logger.info(f"Updating database for conversation: {conversation_id}")
     conn = sqlite3.connect(DB_PATH)
     c = conn.cursor()
     c.execute('''INSERT INTO conversations (user_id, conversation_id, message, response)
                  VALUES (?, ?, ?, ?)''', (user_id, conversation_id, message, response))
     conn.commit()
     conn.close()
+    logger.info("Database updated successfully")
 
 async def clear_inactive_conversations():
     while True:
+        logger.info("Clearing inactive conversations")
         current_time = time.time()
         inactive_convos = [conv_id for conv_id, last_time in last_activity.items() 
                            if current_time - last_time > 1800]  # 30 minutes
@@ -162,10 +185,12 @@ async def clear_inactive_conversations():
                 del conversations[conv_id]
             if conv_id in last_activity:
                 del last_activity[conv_id]
+        logger.info(f"Cleared {len(inactive_convos)} inactive conversations")
         await asyncio.sleep(60)  # Check every minute
 
 @app.on_event("startup")
 async def startup_event():
+    logger.info("Starting up the application")
     FastAPICache.init(InMemoryBackend(), prefix="fastapi-cache")
     asyncio.create_task(clear_inactive_conversations())
 
@@ -183,6 +208,7 @@ async def coding_assistant(query: QueryModel, background_tasks: BackgroundTasks,
     - google/gemma-2-27b-it
     Requires API Key authentication via X-API-Key header.
     """
+    logger.info(f"Received coding assistant query: {query.user_query}")
     if query.conversation_id not in conversations:
         conversations[query.conversation_id] = [
             {"role": "system", "content": "You are a helpful assistant proficient in coding tasks. Help the user in understanding and writing code."}
@@ -200,12 +226,14 @@ async def coding_assistant(query: QueryModel, background_tasks: BackgroundTasks,
             full_response += content
             yield content
         background_tasks.add_task(update_db, query.user_id, query.conversation_id, query.user_query, full_response)
+        logger.info(f"Completed coding assistant response for query: {query.user_query}")
 
     return StreamingResponse(process_response(), media_type="text/event-stream")
 
 # New functions for news assistant
 
 def internet_search(query, type = "web", num_results=20):
+    logger.info(f"Performing internet search for query: {query}, type: {type}")
     if type == "web":
         url = "https://api.search.brave.com/res/v1/web/search"
     else:
@@ -221,6 +249,7 @@ def internet_search(query, type = "web", num_results=20):
     response = requests.get(url, headers=headers, params=params)
 
     if response.status_code != 200:
+        logger.error(f"Failed to fetch search results. Status code: {response.status_code}")
         return []
     
     if type == "web":
@@ -240,17 +269,21 @@ def internet_search(query, type = "web", num_results=20):
         }
         processed_results.append(result)
 
+    logger.info(f"Retrieved {len(processed_results)} search results")
     return processed_results[:num_results]
 
 @lru_cache(maxsize=100)
 def cached_internet_search(query: str):
+    logger.info(f"Performing cached internet search for query: {query}")
     return internet_search(query, type = "news")
 
 
 def analyze_news(query):
+    logger.info(f"Analyzing news for query: {query}")
     news_data = cached_internet_search(query)
     
     if not news_data:
+        logger.error("Failed to fetch news data")
         return "Failed to fetch news data.", []
 
     # Prepare the prompt for the AI
@@ -262,6 +295,7 @@ def analyze_news(query):
         {"role": "user", "content": prompt}
     ]
 
+    logger.info("News analysis completed")
     return messages
 
 @app.post("/news-assistant")
@@ -270,15 +304,18 @@ async def news_assistant(query: NewsQueryModel, api_key: str = Depends(verify_ap
     News assistant endpoint that provides summaries and analysis of recent news based on user queries.
     Requires API Key authentication via X-API-Key header.
     """
+    logger.info(f"Received news assistant query: {query.query}")
     messages = analyze_news(query.query)
     
     if not messages:
+        logger.error("Failed to fetch news data")
         raise HTTPException(status_code=500, detail="Failed to fetch news data")
         
     def process_response():
         for content in chat_with_llama_stream(messages, model=query.model_id):
             yield content
-#meta-llama/llama-3-70b-instruct  google/gemini-pro-1.5
+        logger.info(f"Completed news assistant response for query: {query.query}")
+
     return StreamingResponse(process_response(), media_type="text/event-stream")
 
 class SearchQueryModel(BaseModel):
@@ -296,9 +333,11 @@ class SearchQueryModel(BaseModel):
         }
 
 def analyze_search_results(query):
+    logger.info(f"Analyzing search results for query: {query}")
     search_data = internet_search(query, type="web")
     
     if not search_data:
+        logger.error("Failed to fetch search data")
         return "Failed to fetch search data.", []
 
     # Prepare the prompt for the AI
@@ -309,6 +348,7 @@ def analyze_search_results(query):
         {"role": "user", "content": prompt}
     ]
 
+    logger.info("Search results analysis completed")
     return messages
 
 @app.post("/search-assistant")
@@ -317,17 +357,21 @@ async def search_assistant(query: SearchQueryModel, api_key: str = Depends(verif
     Search assistant endpoint that provides summaries and analysis of web search results based on user queries.
     Requires API Key authentication via X-API-Key header.
     """
+    logger.info(f"Received search assistant query: {query.query}")
     messages = analyze_search_results(query.query)
     
     if not messages:
+        logger.error("Failed to fetch search data")
         raise HTTPException(status_code=500, detail="Failed to fetch search data")
         
     def process_response():
         for content in chat_with_llama_stream(messages, model=query.model_id):
             yield content
+        logger.info(f"Completed search assistant response for query: {query.query}")
 
     return StreamingResponse(process_response(), media_type="text/event-stream")
 
 if __name__ == "__main__":
     import uvicorn
+    logger.info("Starting the application")
     uvicorn.run(app, host="0.0.0.0", port=7860)