Update src/streamlit_app.py

src/streamlit_app.py

@@ -3,6 +3,7 @@ import requests
 import pandas as pd
 from together import Together
 import os
+import time
 
 # =============================================================================
 # CONFIGURATION - Using Secrets Management
@@ -51,34 +52,57 @@ def get_netflix_content():
     
     headers = {"xc-token": api_token}
     all_content = []
-    page = 1
-    page_size = 1000  # Maximum allowed by NocoDB
+    offset = 0
+    page_size = 100  # Smaller page size for better reliability
+    max_retries = 3
     
     try:
         while True:
-            offset = (page - 1) * page_size
-            response = requests.get(
-                f"{NOCODB_URL}{endpoint_path}?limit={page_size}&offset={offset}",
-                headers=headers
-            )
+            retries = 0
+            success = False
             
-            if response.status_code == 200:
-                data = response.json()
-                current_page_content = data.get('list', [])
-                
-                if not current_page_content:  # No more records
-                    break
+            while retries < max_retries and not success:
+                try:
+                    response = requests.get(
+                        f"{NOCODB_URL}{endpoint_path}?limit={page_size}&offset={offset}",
+                        headers=headers,
+                        timeout=10  # Add timeout
+                    )
                     
-                all_content.extend(current_page_content)
-                page += 1
+                    if response.status_code == 200:
+                        data = response.json()
+                        current_page_content = data.get('list', [])
+                        success = True
+                    else:
+                        retries += 1
+                        st.warning(f"Retry {retries}/{max_retries} - Failed to fetch data (status: {response.status_code})")
+                        time.sleep(1)  # Add delay between retries
+                        
+                except (requests.exceptions.RequestException, ValueError) as e:
+                    retries += 1
+                    st.warning(f"Retry {retries}/{max_retries} - Error: {str(e)}")
+                    time.sleep(1)
+            
+            if not success:
+                st.error(f"Failed to fetch data at offset {offset} after {max_retries} retries")
+                break
                 
-                # Show progress
-                st.write(f"Fetched {len(all_content)} records...")
-            else:
-                st.error(f"Failed to fetch data: {response.status_code}")
+            if not current_page_content:  # No more records
                 break
                 
-        st.write(f"Total records fetched: {len(all_content)}")
+            all_content.extend(current_page_content)
+            offset += len(current_page_content)
+            
+            # Show progress
+            progress_text = f"Fetched {len(all_content)} records..."
+            st.write(progress_text)
+            
+            # If we got fewer records than the page size, we've reached the end
+            if len(current_page_content) < page_size:
+                break
+        
+        if all_content:
+            st.success(f"Successfully loaded {len(all_content)} titles!")
         return all_content
             
     except Exception as e:
@@ -193,7 +217,27 @@ def main():
     
     # Header
     st.title("🎬 Netflix Content Explorer")
-    st.write("Explore Netflix movies and TV shows with AI-powered insights!")
+    
+    # Sidebar filters
+    st.sidebar.header("🔍 Filter Content")
+    
+    # Required filters section
+    st.sidebar.subheader("Required Filters")
+    st.sidebar.markdown("*Please select at least one filter to start exploring*")
+    
+    # Content Type (Required)
+    content_type = st.sidebar.selectbox(
+        "Content Type (Required)",
+        options=["Select Type...", "All", "Movie", "TV Show"],
+        index=0
+    )
+    
+    # Initialize filters_selected in session state if not exists
+    if 'filters_selected' not in st.session_state:
+        st.session_state.filters_selected = False
+    
+    # Optional filters section
+    st.sidebar.subheader("Optional Filters")
     
     # Check API credentials
     api_token, together_key, endpoint_path = get_api_credentials()
@@ -208,21 +252,15 @@ def main():
         st.info("Please set NOCODB_API_TOKEN and NOCODB_ENDPOINT_PATH in the Hugging Face Spaces secrets.")
         st.stop()
     
-    # Initialize AI client
-    try:
-        client = get_ai_client()
-        if not client:
-            st.stop()
-    except Exception as e:
-        st.error(f"Failed to initialize Together AI client: {e}")
-        st.stop()
+    # Only fetch data if no content is loaded yet and we need it for filter options
+    if 'all_content' not in st.session_state:
+        with st.spinner("Loading Netflix content..."):
+            st.session_state.all_content = get_netflix_content()
     
-    # Load all content first
-    with st.spinner("Loading Netflix content..."):
-        all_content = get_netflix_content()
+    all_content = st.session_state.all_content
     
     if not all_content:
-        st.error("Could not load Netflix content. Please check your NocoDB connection.")
+        st.error("Could not load Netflix content. Please check your database connection.")
         st.stop()
     
     # Extract unique values for filters
@@ -234,15 +272,7 @@ def main():
         if genre.strip()
     )))
     
-    # Sidebar filters
-    st.sidebar.header("🔍 Filter Content")
-    
-    content_type = st.sidebar.selectbox(
-        "Content Type",
-        options=["All", "Movie", "TV Show"],
-        index=0
-    )
-    
+    # Optional filters
     selected_ratings = st.sidebar.multiselect(
         "Ratings",
         options=all_ratings,
@@ -266,197 +296,260 @@ def main():
     )
     
     # Duration range slider (for movies only)
-    movie_durations = [
-        int(c.get('duration', '0 min').split()[0])
-        for c in all_content
-        if c.get('type') == 'Movie' and 'min' in c.get('duration', '')
-    ]
-    if movie_durations:
-        min_duration, max_duration = min(movie_durations), max(movie_durations)
-        duration_range = st.sidebar.slider(
-            "Movie Duration (minutes)",
-            min_value=min_duration,
-            max_value=max_duration,
-            value=(min_duration, max_duration),
-            help="This filter only applies to movies"
-        )
+    if content_type in ["All", "Movie"]:
+        movie_durations = [
+            int(c.get('duration', '0 min').split()[0])
+            for c in all_content
+            if c.get('type') == 'Movie' and 'min' in c.get('duration', '')
+        ]
+        if movie_durations:
+            min_duration, max_duration = min(movie_durations), max(movie_durations)
+            duration_range = st.sidebar.slider(
+                "Movie Duration (minutes)",
+                min_value=min_duration,
+                max_value=max_duration,
+                value=(min_duration, max_duration),
+                help="This filter only applies to movies"
+            )
+        else:
+            duration_range = (0, 1000)  # Fallback values
     else:
-        duration_range = (0, 1000)  # Fallback values
+        duration_range = (0, 1000)
     
-    # Create filter dictionary
-    filters = {
-        'content_type': content_type if content_type != "All" else None,
-        'ratings': selected_ratings,
-        'genres': selected_genres,
-        'year_range': year_range,
-        'duration_range': duration_range
-    }
+    # Apply button
+    if st.sidebar.button("Apply Filters", type="primary"):
+        if content_type == "Select Type...":
+            st.sidebar.error("Please select a Content Type to continue")
+        else:
+            st.session_state.filters_selected = True
     
-    # Apply filters
-    filtered_content = filter_content(all_content, filters)
+    # Reset button
+    if st.sidebar.button("Reset Filters"):
+        st.session_state.filters_selected = False
+        st.experimental_rerun()
     
     # Main content area
-    col1, col2 = st.columns([2, 1])
-    
-    with col1:
-        # Content listings
-        st.subheader(f"📋 Found {len(filtered_content)} Titles")
+    if not st.session_state.filters_selected:
+        # Landing page content
+        st.markdown("""
+        ## Welcome to Netflix Content Explorer! 🍿
         
-        if filtered_content:
-            # Add pagination controls
-            items_per_page = st.select_slider(
-                "Items per page",
-                options=[10, 20, 50, 100],
-                value=20
-            )
-            
-            total_pages = len(filtered_content) // items_per_page + (1 if len(filtered_content) % items_per_page > 0 else 0)
-            current_page = st.number_input(
-                "Page",
-                min_value=1,
-                max_value=total_pages,
-                value=1
-            )
-            
-            start_idx = (current_page - 1) * items_per_page
-            end_idx = min(start_idx + items_per_page, len(filtered_content))
-            
-            st.write(f"Showing {start_idx + 1} to {end_idx} of {len(filtered_content)} titles")
-            
-            # Show items for current page
-            for i, content in enumerate(filtered_content[start_idx:end_idx], start=start_idx + 1):
-                with st.expander(f"{i}. {content.get('title', 'N/A')} ({content.get('release_year', 'N/A')})"):
-                    # Content details in columns
-                    detail_col1, detail_col2 = st.columns(2)
-                    
-                    with detail_col1:
-                        st.write(f"**🎭 Type:** {content.get('type', 'N/A')}")
-                        st.write(f"**⭐ Rating:** {content.get('rating', 'N/A')}")
-                        st.write(f"**⏱️ Duration:** {content.get('duration', 'N/A')}")
-                    
-                    with detail_col2:
-                        st.write(f"**🎬 Genre:** {content.get('listed_in', 'N/A')}")
-                        # Fix cast handling to prevent NoneType error
-                        cast = content.get('cast')
-                        cast_display = cast[:100] + "..." if cast else "N/A"
-                        st.write(f"**👥 Cast:** {cast_display}")
-                        st.write(f"**📝 Director:** {content.get('director', 'N/A')}")
-                    
-                    # Description
-                    st.write(f"**📖 Description:**")
-                    st.write(content.get('description', 'N/A'))
-            
-            # Add page navigation buttons
-            col1, col2, col3 = st.columns([1, 2, 1])
-            with col1:
-                if current_page > 1:
-                    if st.button("⬅️ Previous"):
-                        st.session_state.page = current_page - 1
-            with col3:
-                if current_page < total_pages:
-                    if st.button("Next ➡️"):
-                        st.session_state.page = current_page + 1
-            
-            # Show page info
-            with col2:
-                st.write(f"Page {current_page} of {total_pages}")
-        else:
-            st.info("No content matches your current filters. Try adjusting the criteria.")
-    
-    with col2:
-        # AI Chat Section
-        st.subheader("🤖 Ask AI Assistant")
-        st.write("Ask questions about the Netflix content!")
+        This application helps you discover and explore Netflix's vast library of movies and TV shows.
         
-        # Model selection for Together AI
-        model_choice = st.selectbox(
-            "Select AI Model:",
-            [
-                "google/gemma-2b-it",
-                "google/gemma-2-27b-it",
-                "mistralai/Mistral-7B-Instruct-v0.1",
-                "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
-                "mistralai/Mixtral-8x7B-Instruct-v0.1"
-            ],
-            help="Select an AI model to answer your questions"
-        )
+        ### How to Get Started:
+        1. **Select Content Type** (Required)
+           - Choose between Movies, TV Shows, or both
         
-        # Example questions
-        with st.expander("💡 Example Questions"):
-            st.write("• What are the most common genres?")
-            st.write("• What's the average movie duration?")
-            st.write("• Which directors have the most content?")
-            st.write("• What are the trends in content ratings?")
-            st.write("• Tell me about the release year distribution")
+        2. **Optional Filters**
+           - Ratings: Filter by content ratings
+           - Genres: Find specific types of content
+           - Release Year: Narrow down by time period
+           - Duration: Filter movies by length
         
-        user_question = st.text_area(
-            "Your Question:",
-            placeholder="Ask about genres, ratings, directors, trends...",
-            height=100
-        )
+        3. **Click 'Apply Filters'** to start exploring!
         
-        if st.button("🔍 Ask AI", type="primary"):
-            if user_question:
-                with st.spinner("AI is analyzing the content..."):
-                    # Create context from current filtered data
-                    context = create_content_context(filtered_content)
-                    
-                    try:
-                        # Get AI response
-                        ai_response = get_ai_response(client, user_question, context, model_choice)
+        ### Features:
+        - 🎯 Advanced filtering capabilities
+        - 🤖 AI-powered insights and recommendations
+        - 📊 Detailed content information
+        - 🔍 Easy-to-use search and navigation
+        
+        ### Dataset Overview:
+        - Total Titles: {total_titles}
+        - Movies: {movies}
+        - TV Shows: {shows}
+        - Year Range: {min_year} - {max_year}
+        - Number of Genres: {genre_count}
+        """.format(
+            total_titles=len(all_content),
+            movies=len([c for c in all_content if c.get('type') == 'Movie']),
+            shows=len([c for c in all_content if c.get('type') == 'TV Show']),
+            min_year=min_year,
+            max_year=max_year,
+            genre_count=len(all_genres)
+        ))
+        
+        # Show some random popular titles as examples
+        st.markdown("### 🌟 Featured Titles")
+        popular_titles = pd.DataFrame(all_content).sample(n=5).to_dict('records')
+        for title in popular_titles:
+            st.markdown(f"- **{title.get('title')}** ({title.get('type')}, {title.get('release_year')})")
+        
+    else:
+        # Create filter dictionary
+        filters = {
+            'content_type': content_type if content_type != "All" else None,
+            'ratings': selected_ratings,
+            'genres': selected_genres,
+            'year_range': year_range,
+            'duration_range': duration_range
+        }
+        
+        # Apply filters
+        filtered_content = filter_content(all_content, filters)
+        
+        # Content display section
+        col1, col2 = st.columns([2, 1])
+        
+        with col1:
+            # Content listings
+            st.subheader(f"📋 Found {len(filtered_content)} Titles")
+            
+            if filtered_content:
+                # Add pagination controls
+                items_per_page = st.select_slider(
+                    "Items per page",
+                    options=[10, 20, 50, 100],
+                    value=20
+                )
+                
+                total_pages = len(filtered_content) // items_per_page + (1 if len(filtered_content) % items_per_page > 0 else 0)
+                current_page = st.number_input(
+                    "Page",
+                    min_value=1,
+                    max_value=total_pages,
+                    value=1
+                )
+                
+                start_idx = (current_page - 1) * items_per_page
+                end_idx = min(start_idx + items_page, len(filtered_content))
+                
+                st.write(f"Showing {start_idx + 1} to {end_idx} of {len(filtered_content)} titles")
+                
+                # Show items for current page
+                for i, content in enumerate(filtered_content[start_idx:end_idx], start=start_idx + 1):
+                    with st.expander(f"{i}. {content.get('title', 'N/A')} ({content.get('release_year', 'N/A')})"):
+                        # Content details in columns
+                        detail_col1, detail_col2 = st.columns(2)
                         
-                        st.success("**AI Assistant Response:**")
-                        st.write(ai_response)
+                        with detail_col1:
+                            st.write(f"**🎭 Type:** {content.get('type', 'N/A')}")
+                            st.write(f"**⭐ Rating:** {content.get('rating', 'N/A')}")
+                            st.write(f"**⏱️ Duration:** {content.get('duration', 'N/A')}")
                         
-                        # Show debug info
-                        with st.expander("Debug Info"):
-                            st.write(f"Model used: {model_choice}")
-                            st.write(f"Content items analyzed: {len(filtered_content)}")
-                            st.write(f"Context: {context[:150]}...")
+                        with detail_col2:
+                            st.write(f"**🎬 Genre:** {content.get('listed_in', 'N/A')}")
+                            # Fix cast handling to prevent NoneType error
+                            cast = content.get('cast')
+                            cast_display = cast[:100] + "..." if cast else "N/A"
+                            st.write(f"**👥 Cast:** {cast_display}")
+                            st.write(f"**📝 Director:** {content.get('director', 'N/A')}")
                         
-                    except Exception as e:
-                        st.error(f"AI Error: {str(e)}")
+                        # Description
+                        st.write(f"**📖 Description:**")
+                        st.write(content.get('description', 'N/A'))
+                
+                # Add page navigation buttons
+                col1, col2, col3 = st.columns([1, 2, 1])
+                with col1:
+                    if current_page > 1:
+                        if st.button("⬅️ Previous"):
+                            st.session_state.page = current_page - 1
+                with col3:
+                    if current_page < total_pages:
+                        if st.button("Next ➡️"):
+                            st.session_state.page = current_page + 1
+                
+                # Show page info
+                with col2:
+                    st.write(f"Page {current_page} of {total_pages}")
+            else:
+                st.info("No content matches your current filters. Try adjusting the criteria.")
+        
+        with col2:
+            # AI Chat Section
+            st.subheader("🤖 Ask AI Assistant")
+            st.write("Ask questions about the Netflix content!")
+            
+            # Model selection for Together AI
+            model_choice = st.selectbox(
+                "Select AI Model:",
+                [
+                    "google/gemma-2b-it",
+                    "google/gemma-2-27b-it",
+                    "mistralai/Mistral-7B-Instruct-v0.1",
+                    "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
+                    "mistralai/Mixtral-8x7B-Instruct-v0.1"
+                ],
+                help="Select an AI model to answer your questions"
+            )
+            
+            # Example questions
+            with st.expander("💡 Example Questions"):
+                st.write("• What are the most common genres?")
+                st.write("• What's the average movie duration?")
+                st.write("• Which directors have the most content?")
+                st.write("• What are the trends in content ratings?")
+                st.write("• Tell me about the release year distribution")
+            
+            user_question = st.text_area(
+                "Your Question:",
+                placeholder="Ask about genres, ratings, directors, trends...",
+                height=100
+            )
+            
+            if st.button("🔍 Ask AI", type="primary"):
+                if user_question:
+                    with st.spinner("AI is analyzing the content..."):
+                        # Create context from current filtered data
+                        context = create_content_context(filtered_content)
                         
-                        # Fallback response with data analysis
-                        st.info("**Fallback Analysis:**")
-                        if filtered_content:
-                            movies = sum(1 for c in filtered_content if c.get('type') == 'Movie')
-                            shows = sum(1 for c in filtered_content if c.get('type') == 'TV Show')
-                            st.write(f"• Found {len(filtered_content)} titles ({movies} movies, {shows} TV shows)")
+                        try:
+                            # Get AI response
+                            ai_response = get_ai_response(client, user_question, context, model_choice)
                             
-                            # Show top genres
-                            genres = [g.strip() for c in filtered_content for g in c.get('listed_in', '').split(',')]
-                            genre_counts = pd.Series(genres).value_counts()
-                            st.write(f"• Top genres: {', '.join(genre_counts.head(3).index)}")
+                            st.success("**AI Assistant Response:**")
+                            st.write(ai_response)
                             
-                            # Show year range
-                            years = [int(c.get('release_year', 0)) for c in filtered_content if c.get('release_year')]
-                            if years:
-                                st.write(f"• Release years: {min(years)} - {max(years)}")
-            else:
-                st.warning("Please enter a question first!")
-    
-    # Footer stats
-    st.markdown("---")
-    if all_content:
-        total_items = len(all_content)
-        filtered_items = len(filtered_content)
-        
-        stat_col1, stat_col2, stat_col3, stat_col4 = st.columns(4)
-        
-        with stat_col1:
-            st.metric("Total Titles", total_items)
-        
-        with stat_col2:
-            st.metric("Filtered Results", filtered_items)
-        
-        with stat_col3:
-            movies = sum(1 for c in filtered_content if c.get('type') == 'Movie')
-            st.metric("Movies", movies)
+                            # Show debug info
+                            with st.expander("Debug Info"):
+                                st.write(f"Model used: {model_choice}")
+                                st.write(f"Content items analyzed: {len(filtered_content)}")
+                                st.write(f"Context: {context[:150]}...")
+                            
+                        except Exception as e:
+                            st.error(f"AI Error: {str(e)}")
+                            
+                            # Fallback response with data analysis
+                            st.info("**Fallback Analysis:**")
+                            if filtered_content:
+                                movies = sum(1 for c in filtered_content if c.get('type') == 'Movie')
+                                shows = sum(1 for c in filtered_content if c.get('type') == 'TV Show')
+                                st.write(f"• Found {len(filtered_content)} titles ({movies} movies, {shows} TV shows)")
+                                
+                                # Show top genres
+                                genres = [g.strip() for c in filtered_content for g in c.get('listed_in', '').split(',')]
+                                genre_counts = pd.Series(genres).value_counts()
+                                st.write(f"• Top genres: {', '.join(genre_counts.head(3).index)}")
+                                
+                                # Show year range
+                                years = [int(c.get('release_year', 0)) for c in filtered_content if c.get('release_year')]
+                                if years:
+                                    st.write(f"• Release years: {min(years)} - {max(years)}")
+                else:
+                    st.warning("Please enter a question first!")
         
-        with stat_col4:
-            shows = sum(1 for c in filtered_content if c.get('type') == 'TV Show')
-            st.metric("TV Shows", shows)
+        # Footer stats
+        st.markdown("---")
+        if all_content:
+            total_items = len(all_content)
+            filtered_items = len(filtered_content)
+            
+            stat_col1, stat_col2, stat_col3, stat_col4 = st.columns(4)
+            
+            with stat_col1:
+                st.metric("Total Titles", total_items)
+            
+            with stat_col2:
+                st.metric("Filtered Results", filtered_items)
+            
+            with stat_col3:
+                movies = sum(1 for c in filtered_content if c.get('type') == 'Movie')
+                st.metric("Movies", movies)
+            
+            with stat_col4:
+                shows = sum(1 for c in filtered_content if c.get('type') == 'TV Show')
+                st.metric("TV Shows", shows)
 
 if __name__ == "__main__":
     main()