Update app.py

app.py

@@ -1,17 +1,13 @@
 """
-DocuMind - Unified FastAPI + Streamlit App for HuggingFace Spaces
-Combines backend API and frontend UI in a single process
+DocuMind - Streamlit Frontend for HuggingFace Spaces
+Simplified version that works with existing FastAPI backend
 """
 
 import streamlit as st
 import requests
 import json
 import os
-import subprocess
 import time
-import sys
-from threading import Thread
-from pathlib import Path
 
 # ============================================================================
 # STREAMLIT CONFIG (Must be first)
@@ -25,43 +21,49 @@ st.set_page_config(
 )
 
 # ============================================================================
-# BACKEND STARTUP
+# API CONFIGURATION
 # ============================================================================
 
-@st.cache_resource
-def start_backend_server():
-    """Start FastAPI backend in a separate thread"""
-    
-    def run_server():
-        # Import FastAPI backend components
-        from main import app as fastapi_app
-        import uvicorn
-        
-        # Run uvicorn server
-        uvicorn.run(
-            fastapi_app,
-            host="127.0.0.1",
-            port=8000,
-            log_level="warning",
-            access_log=False
-        )
-    
-    # Start backend in daemon thread
-    backend_thread = Thread(target=run_server, daemon=True)
-    backend_thread.start()
-    
-    # Give backend time to start
-    time.sleep(2)
-    
-    return "Backend started"
+# Use localhost for internal communication
+API_URL = "http://127.0.0.1:8000"
+API_TIMEOUT = 30  # seconds
+
+# ============================================================================
+# UTILITY FUNCTIONS
+# ============================================================================
+
+def check_backend_health(retries=5):
+    """Check if backend API is running"""
+    for attempt in range(retries):
+        try:
+            response = requests.get(f"{API_URL}/docs", timeout=5)
+            if response.status_code == 200:
+                return True
+        except:
+            if attempt < retries - 1:
+                time.sleep(1)
+    return False
 
-# Start backend
-try:
-    start_backend_server()
-    API_URL = "http://127.0.0.1:8000"
-except Exception as e:
-    st.error(f"Failed to start backend: {e}")
-    st.stop()
+def safe_api_call(method, endpoint, **kwargs):
+    """Safely call API with error handling"""
+    try:
+        url = f"{API_URL}{endpoint}"
+        kwargs.setdefault('timeout', API_TIMEOUT)
+        
+        if method == "GET":
+            response = requests.get(url, **kwargs)
+        elif method == "POST":
+            response = requests.post(url, **kwargs)
+        else:
+            return None, "Invalid method"
+        
+        return response, None
+    except requests.exceptions.Timeout:
+        return None, "⏱️ Request timed out"
+    except requests.exceptions.ConnectionError:
+        return None, "⚠️ Backend not responding"
+    except Exception as e:
+        return None, f"❌ Error: {str(e)}"
 
 # ============================================================================
 # STREAMLIT FRONTEND UI
@@ -71,6 +73,24 @@ st.title("🧠 DocuMind")
 st.markdown("**Enterprise Document Intelligence Chatbot**")
 st.markdown("---")
 
+# Health check on load
+if "backend_checked" not in st.session_state:
+    st.session_state.backend_checked = False
+    st.session_state.backend_healthy = False
+
+if not st.session_state.backend_checked:
+    with st.spinner("🔄 Starting backend..."):
+        time.sleep(2)  # Give backend time to start
+        st.session_state.backend_healthy = check_backend_health()
+    st.session_state.backend_checked = True
+
+# Show status
+if not st.session_state.backend_healthy:
+    st.warning(
+        "⚠️ Backend is starting up. This may take 30-60 seconds on first load. "
+        "Please refresh the page if you see this message for more than 1 minute."
+    )
+
 # --- Sidebar for Document Upload ---
 with st.sidebar:
     st.header("🏢 Document Knowledge Base")
@@ -81,35 +101,31 @@ with st.sidebar:
     if uploaded_file and st.button("Ingest Document", key="ingest_btn"):
         with st.spinner("Ingesting document (creating chunks & embeddings)..."):
             files = {"file": (uploaded_file.name, uploaded_file.getvalue())}
-            try:
-                res = requests.post(f"{API_URL}/ingest", files=files, timeout=30)
-                if res.status_code == 200:
-                    st.success(f"✅ {uploaded_file.name} ingested successfully!")
-                else:
-                    st.error(f"❌ Failed to ingest: {res.text}")
-            except requests.exceptions.ConnectionError:
-                st.error("⚠️ Backend is not running. Please refresh the page.")
-            except Exception as e:
-                st.error(f"⚠️ Error: {str(e)}")
+            response, error = safe_api_call("POST", "/ingest", files=files)
+            
+            if error:
+                st.error(error)
+            elif response and response.status_code == 200:
+                st.success(f"✅ {uploaded_file.name} ingested successfully!")
+            else:
+                st.error(f"❌ Failed to ingest: {response.text if response else 'Unknown error'}")
     
     st.divider()
     
     st.subheader("📄 Indexed Documents")
-    try:
-        res = requests.get(f"{API_URL}/sources", timeout=10)
-        if res.status_code == 200:
-            documents = res.json().get("documents", [])
-            if documents:
-                for doc in documents:
-                    st.markdown(f"- `{doc}`")
-            else:
-                st.info("No documents indexed yet.")
+    response, error = safe_api_call("GET", "/sources")
+    
+    if error:
+        st.warning(f"Could not fetch documents: {error}")
+    elif response and response.status_code == 200:
+        documents = response.json().get("documents", [])
+        if documents:
+            for doc in documents:
+                st.markdown(f"- `{doc}`")
         else:
-            st.warning("Could not fetch documents list")
-    except requests.exceptions.ConnectionError:
-        st.warning("⚠️ Could not connect to backend")
-    except Exception as e:
-        st.warning(f"⚠️ Error: {str(e)}")
+            st.info("No documents indexed yet.")
+    else:
+        st.info("Backend not ready yet...")
 
 # --- Main Chat Interface ---
 st.subheader("💬 Chat with Your Documents")
@@ -128,7 +144,11 @@ for msg in st.session_state.messages:
                     score = src.get('score', 0)
                     st.caption(f"**Source {idx+1}** [Relevance: {score:.2%}]")
                     st.markdown(f"**From:** `{src.get('source', 'Unknown')}`")
-                    st.markdown(f"> {src.get('content', '')[:500]}...")
+                    content = src.get('content', '')
+                    if len(content) > 500:
+                        st.markdown(f"> {content[:500]}...")
+                    else:
+                        st.markdown(f"> {content}")
 
 # Chat input
 user_input = st.chat_input("Ask a question about your documents...")
@@ -147,16 +167,20 @@ if user_input:
         full_response = ""
         sources = []
         
-        try:
-            with requests.post(
-                f"{API_URL}/query",
-                json={"question": user_input},
-                stream=True,
-                timeout=60
-            ) as r:
-                r.raise_for_status()
-                
-                for line in r.iter_lines():
+        response, error = safe_api_call(
+            "POST",
+            "/query",
+            json={"question": user_input},
+            stream=True
+        )
+        
+        if error:
+            error_msg = error
+            st.error(error_msg)
+            full_response = error_msg
+        elif response:
+            try:
+                for line in response.iter_lines():
                     if line:
                         try:
                             decoded_line = line.decode('utf-8')
@@ -179,20 +203,16 @@ if user_input:
                             score = src.get('score', 0)
                             st.caption(f"**Source {idx+1}** [Relevance: {score:.2%}]")
                             st.markdown(f"**From:** `{src.get('source', 'Unknown')}`")
-                            st.markdown(f"> {src.get('content', '')[:500]}...")
-        
-        except requests.exceptions.Timeout:
-            error_msg = "⏱️ Request timed out. Please try a shorter question."
-            st.error(error_msg)
-            full_response = error_msg
-        except requests.exceptions.ConnectionError:
-            error_msg = "⚠️ Backend connection failed. Please refresh the page."
-            st.error(error_msg)
-            full_response = error_msg
-        except Exception as e:
-            error_msg = f"❌ Error: {str(e)}"
-            st.error(error_msg)
-            full_response = error_msg
+                            content = src.get('content', '')
+                            if len(content) > 500:
+                                st.markdown(f"> {content[:500]}...")
+                            else:
+                                st.markdown(f"> {content}")
+            
+            except Exception as e:
+                error_msg = f"❌ Error processing response: {str(e)}"
+                st.error(error_msg)
+                full_response = error_msg
     
     # Save assistant message
     st.session_state.messages.append({
@@ -204,7 +224,7 @@ if user_input:
 # Footer
 st.divider()
 st.markdown(
-    "<div style='text-align: center; color: gray; font-size: 0.8em;'>"
+    "<div style='text-align: center; color: var(--color-text-secondary); font-size: 0.85em;'>"
     "DocuMind - Enterprise RAG Chatbot | "
     "<a href='https://github.com/TejeshNaiduKona/DocuMind' target='_blank'>GitHub</a>"
     "</div>",