Fix import errors in app.py by using provider classes directly instead of legacy functions

app.py

@@ -8,7 +8,6 @@ from datetime import datetime
 from pathlib import Path
 sys.path.append(str(Path(__file__).parent))
 from utils.config import config
-from core.llm import send_to_ollama, send_to_hf
 from core.session import session_manager
 from core.memory import check_redis_health
 from core.coordinator import coordinator
@@ -16,6 +15,9 @@ from core.errors import translate_error
 from core.personality import personality
 from services.hf_endpoint_monitor import hf_monitor
 from services.weather import weather_service
+from core.llm import LLMClient
+from core.providers.ollama import OllamaProvider
+from core.providers.huggingface import HuggingFaceProvider
 import logging
 
 # Set up logging
@@ -78,21 +80,14 @@ with st.sidebar:
         
     if st.button("📡 Test Connection"):
         try:
-            import requests
-            headers = {
-                "ngrok-skip-browser-warning": "true",
-                "User-Agent": "CosmicCat-Test"
-            }
-            with st.spinner("Testing connection..."):
-                response = requests.get(
-                    f"{ngrok_url_input}/api/tags",
-                    headers=headers,
-                    timeout=15
-                )
-                if response.status_code == 200:
-                    st.success("✅ Connection successful!")
-                else:
-                    st.error(f"❌ Failed: {response.status_code}")
+            # Use OllamaProvider to test connection
+            ollama_provider = OllamaProvider(st.session_state.selected_model)
+            # Test model validation
+            is_valid = ollama_provider.validate_model()
+            if is_valid:
+                st.success("✅ Connection successful!")
+            else:
+                st.error("❌ Model validation failed")
         except Exception as e:
             st.error(f"❌ Error: {str(e)[:50]}...")
             
@@ -282,10 +277,11 @@ if user_input and not st.session_state.is_processing:
             response_placeholder = st.empty()
             
             try:
-                # Get conversation history
+                # Get conversation history from session
                 user_session = session_manager.get_session("default_user")
-                conversation = user_session.get("conversation", [])
-                conversation_history = conversation[-5:]  # Last 5 messages
+                conversation_history = user_session.get("conversation", []).copy()
+                
+                # Add the current user message to history for context
                 conversation_history.append({"role": "user", "content": validated_input})
                 
                 # Check if cosmic mode is enabled
@@ -297,17 +293,14 @@ if user_input and not st.session_state.is_processing:
                         # Get conversation history
                         user_session = session_manager.get_session("default_user")
                         conversation_history = user_session.get("conversation", []).copy()
-
+                        conversation_history.append({"role": "user", "content": validated_input})
+                        
                         # Stage 1: Local Ollama Response
-                        local_response = send_to_ollama(
-                            validated_input,
-                            conversation_history,
-                            st.session_state.ngrok_url_temp,
-                            st.session_state.selected_model
-                        )
-
+                        ollama_provider = OllamaProvider(st.session_state.selected_model)
+                        local_response = ollama_provider.generate(validated_input, conversation_history)
+                        
                         if local_response:
-                            # Display response (no nested chat_message)
+                            # Display response
                             st.markdown(f"### 🐱 Cosmic Kitten Says:\n{local_response}")
                             st.session_state.messages.append({
                                 "role": "assistant",
@@ -315,7 +308,7 @@ if user_input and not st.session_state.is_processing:
                                 "source": "local_kitty",
                                 "timestamp": datetime.now().strftime("%H:%M:%S")
                             })
-                            
+                        
                         # Stage 2: HF Endpoint Analysis
                         status_placeholder.info("🛰️ Beaming Query to Orbital Station...")
                         if config.hf_token:
@@ -323,11 +316,12 @@ if user_input and not st.session_state.is_processing:
                             hf_status = hf_monitor.check_endpoint_status()
                             if not hf_status['available']:
                                 status_placeholder.info(personality.get_initializing_message())
-                                
-                            hf_response = send_to_hf(validated_input, conversation_history)
+                            
+                            hf_provider = HuggingFaceProvider("meta-llama/Llama-2-7b-chat-hf")
+                            hf_response = hf_provider.generate(validated_input, conversation_history)
                             
                             if hf_response:
-                                # Display response (no nested chat_message)
+                                # Display response
                                 st.markdown(f"### 🛰️ Orbital Station Reports:\n{hf_response}")
                                 st.session_state.messages.append({
                                     "role": "assistant",
@@ -335,7 +329,7 @@ if user_input and not st.session_state.is_processing:
                                     "source": "orbital_station",
                                     "timestamp": datetime.now().strftime("%H:%M:%S")
                                 })
-                                
+                        
                         # Stage 3: Local Synthesis
                         status_placeholder.info("🐱 Cosmic Kitten Synthesizing Wisdom...")
                         
@@ -346,15 +340,13 @@ if user_input and not st.session_state.is_processing:
                             {"role": "assistant", "content": hf_response, "source": "cloud"}
                         ])
                         
-                        synthesis = send_to_ollama(
+                        synthesis = ollama_provider.generate(
                             f"Synthesize these two perspectives:\n1. Local: {local_response}\n2. Cloud: {hf_response}",
-                            synthesis_history,
-                            st.session_state.ngrok_url_temp,
-                            st.session_state.selected_model
+                            synthesis_history
                         )
                         
                         if synthesis:
-                            # Display response (no nested chat_message)
+                            # Display response
                             st.markdown(f"### 🌟 Final Cosmic Summary:\n{synthesis}")
                             st.session_state.messages.append({
                                 "role": "assistant",
@@ -362,7 +354,7 @@ if user_input and not st.session_state.is_processing:
                                 "source": "cosmic_summary",
                                 "timestamp": datetime.now().strftime("%H:%M:%S")
                             })
-                            
+                        
                         status_placeholder.success("✨ Cosmic Cascade Complete!")
                         
                     except Exception as e:
@@ -376,17 +368,14 @@ if user_input and not st.session_state.is_processing:
                         })
                 else:
                     # Traditional processing
-                    # Try Ollama with proper error handling
+                    # Try Ollama first
                     status_placeholder.info("🦙 Contacting Ollama...")
                     ai_response = None
                     
                     try:
-                        ai_response = send_to_ollama(
-                            validated_input,
-                            conversation_history,
-                            st.session_state.ngrok_url_temp,
-                            st.session_state.selected_model
-                        )
+                        # Use the OllamaProvider directly
+                        ollama_provider = OllamaProvider(st.session_state.selected_model)
+                        ai_response = ollama_provider.generate(validated_input, conversation_history)
                         
                         if ai_response:
                             st.markdown(ai_response)  # Use st.markdown instead of response_placeholder
@@ -408,7 +397,9 @@ if user_input and not st.session_state.is_processing:
                                 if not hf_status['available']:
                                     status_placeholder.info(personality.get_initializing_message())
                                     
-                                ai_response = send_to_hf(validated_input, conversation_history)
+                                # Use the HuggingFaceProvider directly
+                                hf_provider = HuggingFaceProvider("meta-llama/Llama-2-7b-chat-hf")
+                                ai_response = hf_provider.generate(validated_input, conversation_history)
                                 
                                 if ai_response:
                                     st.markdown(ai_response)  # Use st.markdown instead of response_placeholder
@@ -423,6 +414,7 @@ if user_input and not st.session_state.is_processing:
                     # Save response if successful
                     if ai_response:
                         # Update conversation history
+                        conversation = user_session.get("conversation", [])
                         conversation.append({"role": "user", "content": validated_input})
                         conversation.append({"role": "assistant", "content": ai_response})
                         user_session["conversation"] = conversation