Update src/streamlit_app.py

src/streamlit_app.py

@@ -1,853 +1,167 @@
-import streamlit as st
-import requests
-import pandas as pd
-from together import Together
-import os
-import json
-from collections import defaultdict
-import re
-
-# =============================================================================
-# CONFIGURATION - Using Secrets Management
-# =============================================================================
-NOCODB_URL = "https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/mvtt3arw5ni7uqp"  # Updated with table ID
-
-# Get sensitive data from Streamlit secrets or environment variables
-def get_api_credentials():
-    """Get API credentials from secrets or environment"""
-    try:
-        # Try Streamlit secrets first (for Hugging Face Spaces)
-        api_token = st.secrets.get("NOCODB_API_TOKEN", os.environ.get("NOCODB_API_TOKEN", ""))
-        together_key = st.secrets.get("TOGETHER_API_KEY", os.environ.get("TOGETHER_API_KEY", ""))
-        
-        return api_token, together_key
-    except:
-        # Fallback to environment variables
-        api_token = os.environ.get("NOCODB_API_TOKEN", "")
-        together_key = os.environ.get("TOGETHER_API_KEY", "")
-        
-        return api_token, together_key
-
-# Initialize Together AI client
-@st.cache_resource
-def get_ai_client():
-    """Initialize Together AI client"""
-    _, together_key = get_api_credentials()
-    if not together_key:
-        st.error("Together AI API key not found. Please configure it in the secrets.")
-        return None
-    return Together(api_key=together_key)
-
-# =============================================================================
-# HELPER FUNCTIONS
-# =============================================================================
-@st.cache_data(ttl=300)  # Cache for 5 minutes
-def get_streaming_content():
-    """Fetch streaming content from NocoDB with pagination"""
-    api_token, _ = get_api_credentials()
-    
-    if not api_token:
-        st.error("NocoDB credentials not configured. Please set up your secrets.")
-        return []
-    
-    headers = {
-        "xc-token": api_token,
-        "accept": "application/json"
-    }
-    
-    all_content = []
-    page = 1
-    page_size = 1000  # NocoDB default page size
-    
-    try:
-        while True:
-            offset = (page - 1) * page_size
-            url = f"{NOCODB_URL}?limit={page_size}&offset={offset}"
-            
-            response = requests.get(url, headers=headers)
-            
-            if response.status_code == 200:
-                data = response.json()
-                current_page_data = data.get('list', [])
-                
-                # Filter out None values and ensure all items are dictionaries
-                current_page_data = [item for item in current_page_data if item and isinstance(item, dict)]
-                
-                if not current_page_data:  # No more data to fetch
-                    break
-                    
-                all_content.extend(current_page_data)
-                
-                # Check if this is the last page
-                page_info = data.get('pageInfo', {})
-                if page_info.get('isLastPage', True):
-                    break
-                
-                page += 1
-            else:
-                st.error(f"Failed to fetch data: {response.status_code}")
-                if not all_content:  # Only return [] if we haven't fetched any data
-                    return []
-                break  # If we have some data, return what we've got
-                
-        return all_content
-            
-    except Exception as e:
-        st.error(f"Error connecting to database: {str(e)}")
-        st.write("Full error details:", e)
-        return []
-
-def filter_content(content_list, filters):
-    """Apply filters to streaming content list"""
-    filtered = []
-    
-    for content in content_list:
-        if not content or not isinstance(content, dict):
-            continue
-            
-        matches_all_filters = True
-        
-        # Streaming service filter
-        if filters['streaming_services']:
-            if content.get('streaming_service') not in filters['streaming_services']:
-                matches_all_filters = False
-                continue
-        
-        # Type filter - only apply if not "All"
-        if filters['content_type']:
-            if content.get('type') != filters['content_type']:
-                matches_all_filters = False
-                continue
-            
-        # Genre filter - check if ALL selected genres are in the content's genres
-        if filters['genres']:
-            content_genres = set(g.strip().lower() for g in str(content.get('listed_in', '')).split(','))
-            selected_genres = set(g.strip().lower() for g in filters['genres'])
-            
-            if not selected_genres.issubset(content_genres):
-                matches_all_filters = False
-                continue
-            
-        # Rating filter
-        if filters['ratings']:
-            rating = content.get('rating', '').strip()
-            # Only compare if rating is a valid string and not a duration
-            if not rating or not isinstance(rating, str) or rating.endswith('min'):
-                matches_all_filters = False
-                continue
-            if rating not in filters['ratings']:
-                matches_all_filters = False
-                continue
-
-        # Release year filter
-        try:
-            release_year = int(content.get('release_year', 0))
-            if release_year < filters['year_range'][0] or release_year > filters['year_range'][1]:
-                matches_all_filters = False
-                continue
-        except (ValueError, TypeError):
-            matches_all_filters = False
-            continue
-            
-        # Duration filter (different handling for movies)
-        if filters['content_type'] == 'Movie':
-            duration = str(content.get('duration', ''))
-            if 'min' in duration:
-                try:
-                    minutes = int(duration.split()[0])
-                    if minutes < filters['duration_range'][0] or minutes > filters['duration_range'][1]:
-                        matches_all_filters = False
-                        continue
-                except (ValueError, IndexError):
-                    matches_all_filters = False
-                    continue
-
-        # Director filter (optional)
-        if filters['director']:
-            director = str(content.get('director', '')).lower()
-            if not any(name.strip().lower() in director for name in filters['director'].split(',')):
-                matches_all_filters = False
-                continue
-
-        # Cast filter (optional)
-        if filters['cast']:
-            cast = str(content.get('cast', '')).lower()
-            if not any(name.strip().lower() in cast for name in filters['cast'].split(',')):
-                matches_all_filters = False
-                continue
-        
-        if matches_all_filters:
-            filtered.append(content)
-    
-    return filtered
-
-def create_content_context(content_list):
-    """Create context string about current content for AI"""
-    if not content_list:
-        return "No content matches the current filters."
-    
-    total = len(content_list)
-    movies = sum(1 for c in content_list if c.get('type') == 'Movie')
-    shows = sum(1 for c in content_list if c.get('type') == 'TV Show')
-    
-    context = f"""Currently showing {total} titles ({movies} movies and {shows} TV shows). """
-    
-    # Add some genre info
-    all_genres = []
-    for content in content_list[:20]:  # Sample from first 20 items
-        genres = content.get('listed_in', '').split(', ')
-        all_genres.extend(genres)
-    
-    if all_genres:
-        genre_counts = pd.Series(all_genres).value_counts()
-        top_genres = genre_counts.head(5).index.tolist()
-        context += f"Top genres include: {', '.join(top_genres)}. "
-    
-    # Add year range
-    years = [int(c.get('release_year', 0)) for c in content_list if c.get('release_year')]
-    if years:
-        context += f"Release years range from {min(years)} to {max(years)}."
-    
-    return context
-
-def get_ai_response(client, question, context, model_name):
-    """Get response from Together AI"""
-    try:
-        prompt = f"""You are a helpful streaming content expert. Based on the current content data, please answer the user's question accurately and helpfully.
-
-Current Content Data Context:
-{context}
-
-User Question: {question}
-
-Please provide a helpful, accurate response based on the data provided. Keep your answer concise but informative."""
-
-        response = client.chat.completions.create(
-            model=model_name,
-            messages=[
-                {"role": "system", "content": "You are a helpful content expert with deep knowledge of movies and TV shows."},
-                {"role": "user", "content": prompt}
-            ],
-            max_tokens=300,
-            temperature=0.7,
-        )
-        
-        return response.choices[0].message.content
-        
-    except Exception as e:
-        raise Exception(f"Together AI Error: {str(e)}")
-
-def extract_unique_names(content_list, field):
-    """Extract unique names from a comma-separated field in content list"""
-    unique_names = set()
-    for content in content_list:
-        names = content.get(field, '')
-        if names:
-            # Split by comma and clean each name
-            for name in names.split(','):
-                cleaned_name = name.strip()
-                if cleaned_name:  # Only add non-empty names
-                    unique_names.add(cleaned_name)
-    return sorted(list(unique_names))
-
-def get_similar_content(content, n_recommendations=5):
-    """Get pre-computed similar content from database"""
-    try:
-        # Get database credentials
-        api_token, _ = get_api_credentials()
-        headers = {
-            "xc-token": api_token,
-            "accept": "application/json"
-        }
-        
-        # Debug: Print the title and show_id we're searching for
-        title = content.get('title', '')
-        show_id = content.get('show_id', '')
-        st.write(f"🔍 Searching for similar content to: {title} (ID: {show_id})")
-        
-        # Base URL for the NocoDB API
-        base_url = "https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee"
-        
-        # URLs for both similarity tables using table IDs
-        similarity_table_urls = [
-            f"{base_url}/mp7bnn9tzhojh7k",  # Part 1 similarities
-            f"{base_url}/m8e5rglns4acmef",   # Part 2 similarities
-            f"{base_url}/m2driodimid10k6"    # Part 3 similarities
-        ]
-        
-        similar_items = []
-        
-        # Check both tables for similarities
-        for table_url in similarity_table_urls:
-            # Debug: Get a sample row to see table structure
-            sample_params = {
-                "limit": 1
-            }
-            sample_response = requests.get(table_url, headers=headers, params=sample_params)
-            if sample_response.status_code == 200:
-                sample_data = sample_response.json()
-                if sample_data.get('list'):
-                    st.write(f"📋 Table structure for {table_url.split('/')[-1]}:")
-                    st.write("Columns:", list(sample_data['list'][0].keys()))
-                    st.write("Sample row:", sample_data['list'][0])
-            
-            # Try finding by show_id
-            query = f'(show_id,eq,"{show_id}")'
-            alt_query = f'(show_id,like,"{show_id}")'  # Try alternative query format
-            
-            params = {
-                "where": query
-            }
-            alt_params = {
-                "where": alt_query
-            }
-            
-            # Debug: Print the request details
-            st.write(f"📡 Querying table: {table_url.split('/')[-1]}")
-            st.write(f"🔍 Query params: {params}")
-            
-            # Try first query format
-            try:
-                response = requests.get(table_url, headers=headers, params=params)
-                st.write(f"📥 Response status: {response.status_code}")
-                st.write("📥 Request URL:", response.url)
-                
-                # If first query returns no results, try alternative query
-                if response.status_code == 200:
-                    data = response.json()
-                    if not data.get('list'):
-                        st.write("🔄 Trying alternative query format...")
-                        response = requests.get(table_url, headers=headers, params=alt_params)
-                        st.write(f"📥 Alt response status: {response.status_code}")
-                        st.write("📥 Alt request URL:", response.url)
-                        if response.status_code == 200:
-                            data = response.json()
-                
-                # Debug: Print raw response
-                try:
-                    response_json = response.json()
-                    st.write("Raw response:", response_json)
-                    
-                    # Additional debugging
-                    if response_json.get('list'):
-                        st.write("✅ Found entries in response")
-                        for entry in response_json['list']:
-                            st.write(f"Entry show_id: {entry.get('show_id')}")
-                    else:
-                        st.write("❌ No entries found in response")
-                        
-                        # Try to get a few random entries to verify data
-                        sample_params = {
-                            "limit": 3,
-                            "shuffle": true
-                        }
-                        sample_response = requests.get(table_url, headers=headers, params=sample_params)
-                        if sample_response.status_code == 200:
-                            sample_data = sample_response.json()
-                            st.write("🔍 Random entries from table:")
-                            for entry in sample_data.get('list', []):
-                                st.write(f"- {entry.get('title')} (ID: {entry.get('show_id')})")
-                    
-                except Exception as e:
-                    st.write("Could not parse response as JSON:", response.text)
-                
-                if response.status_code == 200:
-                    data = response_json
-                    st.write(f"📊 Found {len(data.get('list', []))} matches in table")
-                    
-                    if data and len(data.get('list', [])) > 0:
-                        # Get similar items from stored data
-                        try:
-                            table_items = json.loads(data['list'][0]['similar_items'])
-                            st.write(f"✅ Successfully parsed {len(table_items)} similar items")
-                            similar_items.extend(table_items)
-                        except Exception as parse_error:
-                            st.write(f"❌ Error parsing similar items: {str(parse_error)}")
-                            st.write("Raw data:", data['list'][0])
-            except Exception as e:
-                st.write(f"❌ Request error: {str(e)}")
-        
-        st.write(f"📝 Total similar items found: {len(similar_items)}")
-        
-        if similar_items:
-            # Sort by similarity score and limit to requested number
-            similar_items.sort(key=lambda x: x['similarity'], reverse=True)
-            similar_items = similar_items[:n_recommendations]
-            
-            # Get full content details for each similar item
-            similar_content = []
-            for item in similar_items:
-                # Query main content table for full details using show_id if available
-                show_id = item.get('show_id', '')
-                if show_id:
-                    query = f'(show_id,eq,"{show_id}")'
-                else:
-                    # Fallback to title if show_id not available
-                    query = f'(title,eq,"{item["title"]}")'
-                    
-                content_params = {
-                    "where": query
-                }
-                content_response = requests.get(NOCODB_URL, headers=headers, params=content_params)
-                
-                if content_response.status_code == 200:
-                    content_data = content_response.json()
-                    if content_data and len(content_data.get('list', [])) > 0:
-                        content_dict = content_data['list'][0]
-                        content_dict['similarity'] = f"{item['similarity']:.2%}"
-                        similar_content.append(content_dict)
-                        st.write(f"✅ Found details for: {item['title']}")
-                    else:
-                        st.write(f"❌ No content details found for: {item['title']}")
-            
-            return similar_content
-        
-        return []
-    except Exception as e:
-        st.error(f"Error fetching similar content: {str(e)}")
-        st.write("Full error details:", e)
-        return []
-
-# =============================================================================
-# MAIN APP
-# =============================================================================
-def main():
-    # Page config
-    st.set_page_config(
-        page_title="Streaming Content Explorer",
-        page_icon="🎬",
-        layout="wide"
-    )
-    
-    # Header
-    st.title("🎬 Streaming Content Explorer")
-    st.write("Explore movies and TV shows across multiple streaming platforms!")
-    
-    # Check API credentials
-    api_token, together_key = get_api_credentials()
-    
-    if not together_key:
-        st.error("⚠️ Together AI API key not configured!")
-        st.info("Please set your TOGETHER_API_KEY in the Hugging Face Spaces secrets.")
-        st.stop()
-    
-    if not api_token:
-        st.error("⚠️ NocoDB credentials not configured!")
-        st.info("Please set NOCODB_API_TOKEN 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()
-    
-    # Load all content first
-    with st.spinner("Loading streaming content..."):
-        all_content = get_streaming_content()
-    
-    if not all_content:
-        st.error("Could not load streaming content. Please check your NocoDB connection.")
-        st.stop()
-    
-    # Extract unique values for filters
-    all_ratings = sorted(list(set(
-        c.get('rating') for c in all_content 
-        if c and isinstance(c, dict) 
-        and c.get('rating') 
-        and isinstance(c.get('rating'), str)
-        and not c.get('rating').endswith('min')  # Exclude duration values
-        and c.get('rating').strip()  # Exclude empty strings
-    )))
-    all_genres = sorted(list(set(
-        genre.strip()
-        for c in all_content
-        for genre in c.get('listed_in', '').split(',')
-        if genre.strip()
-    )))
-    all_streaming_services = sorted(list(set([c.get('streaming_service') for c in all_content if c.get('streaming_service')])))
-    
-    # Extract unique directors and cast members
-    all_directors = extract_unique_names(all_content, 'director')
-    all_cast_members = extract_unique_names(all_content, 'cast')
-    
-    # Sidebar filters
-    st.sidebar.header("🔍 Filter Content")
-
-    with st.sidebar.form("filter_form"):
-        st.subheader("Streaming Services")
-        
-        # Streaming service selection (required)
-        selected_services = st.multiselect(
-            "Select Your Streaming Services",
-            options=all_streaming_services,
-            default=all_streaming_services[:1],  # Default to first service
-            help="Select the streaming services you have access to",
-            key="streaming_services"
-        )
-        
-        if not selected_services:
-            st.warning("Please select at least one streaming service")
-        
-        st.subheader("Content Filters")
-        
-        content_type = st.selectbox(
-            "Content Type",
-            options=["All", "Movie", "TV Show"],
-            index=0
-        )
-        
-        selected_genres = st.multiselect(
-            "Genres",
-            options=all_genres,
-            default=[]
-        )
-
-        st.subheader("Optional Filters")
-        
-        # Rating filter
-        selected_ratings = st.multiselect(
-            "Ratings",
-            options=all_ratings,
-            default=[],
-            help="Filter by content rating"
-        )
-
-        # Year range slider
-        years = [int(c.get('release_year', 0)) for c in all_content if c.get('release_year')]
-        min_year, max_year = min(years), max(years)
-        year_range = st.slider(
-            "Release Year",
-            min_value=min_year,
-            max_value=max_year,
-            value=(min_year, max_year),
-            help="Filter by release year range"
-        )
-        
-        # Duration range slider (for movies only)
-        movie_durations = [
-            int(str(c.get('duration', '0 min')).split()[0])
-            for c in all_content
-            if c and c.get('type') == 'Movie' and 'min' in str(c.get('duration', ''))
-        ]
-        
-        if movie_durations:
-            min_duration = min(d for d in movie_durations if d > 0)
-            max_duration = max(movie_durations)
-            duration_range = st.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
-
-        # Director filter with autocomplete
-        selected_directors = st.multiselect(
-            "Directors",
-            options=all_directors,
-            default=[],
-            help="Select one or more directors (searchable)",
-            placeholder="Start typing to search directors..."
-        )
-
-        # Cast filter with autocomplete
-        selected_cast = st.multiselect(
-            "Cast Members",
-            options=all_cast_members,
-            default=[],
-            help="Select one or more cast members (searchable)",
-            placeholder="Start typing to search cast members..."
-        )
-        
-        # Submit button
-        apply_filters = st.form_submit_button("🔍 Apply Filters", type="primary")
-    
-    # Create filter dictionary
-    filters = {
-        'streaming_services': selected_services,
-        'content_type': content_type if content_type != "All" else None,
-        'ratings': selected_ratings,
-        'genres': selected_genres,
-        'year_range': year_range,
-        'duration_range': duration_range,
-        'director': ','.join(selected_directors) if selected_directors else '',
-        'cast': ','.join(selected_cast) if selected_cast else ''
-    }
-    
-    # Only apply filters when the button is clicked
-    if apply_filters:
-        filtered_content = filter_content(all_content, filters)
-        st.session_state.filtered_content = filtered_content
-    else:
-        # Initialize filtered content if not exists
-        if 'filtered_content' not in st.session_state:
-            st.session_state.filtered_content = all_content
-
-    # Main content area
-    col1, col2 = st.columns([2, 1])
-    
-    with col1:
-        # Content listings
-        filtered_count = len(st.session_state.filtered_content)
-        if filtered_count == 0:
-            st.subheader("📋 No Titles Found")
-        else:
-            # Header with count and page info
-            st.subheader(f"📋 Found {filtered_count:,} Title{'s' if filtered_count != 1 else ''}")
-        
-        if st.session_state.filtered_content:
-            # Active Filters section with better formatting
-            if any([filters['content_type'], filters['genres'], filters['ratings'], 
-                   filters['director'], filters['cast']]):
-                with st.expander("🔍 Active Filters", expanded=True):
-                    filter_cols = st.columns(2)
-                    with filter_cols[0]:
-                        if filters['content_type']:
-                            st.write(f"**Type:** {filters['content_type']}")
-                        if filters['genres']:
-                            st.write(f"**Genres:** {', '.join(filters['genres'])}")
-                        if filters['ratings']:
-                            st.write(f"**Ratings:** {', '.join(filters['ratings'])}")
-                    with filter_cols[1]:
-                        if filters['director']:
-                            st.write(f"**Director:** {filters['director']}")
-                        if filters['cast']:
-                            st.write(f"**Cast:** {filters['cast']}")
-            st.write("---")
-
-            # Pagination setup
-            items_per_page = 10
-            total_pages = (filtered_count + items_per_page - 1) // items_per_page
-
-            # Initialize page number in session state if not exists
-            if 'current_page' not in st.session_state:
-                st.session_state.current_page = 1
-
-            # Calculate slice indices for current page
-            start_idx = (st.session_state.current_page - 1) * items_per_page
-            end_idx = min(start_idx + items_per_page, filtered_count)
-            
-            # Display current range info
-            st.write(f"Showing {start_idx + 1}-{end_idx} of {filtered_count:,} titles")
-            
-            # Show items for current page
-            for i, content in enumerate(st.session_state.filtered_content[start_idx:end_idx], start=start_idx):
-                with st.expander(f"{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"**📺 Available on:** {content.get('streaming_service', 'N/A')}")
-                        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"**🎬 Genres:** {content.get('listed_in', 'N/A')}")
-                        cast = content.get('cast')
-                        cast_display = cast[:100] + "..." if cast and len(cast) > 100 else cast 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 Find Similar button with loading state
-                    similar_button = st.button(f"🔍 Find Similar Content", key=f"similar_{i}")
-                    if similar_button:
-                        with st.spinner("Finding similar content..."):
-                            similar_content = get_similar_content(content)
-                            
-                            if similar_content:
-                                # Create tabs for different aspects of recommendations
-                                sim_tab1, sim_tab2 = st.tabs(["📺 Similar Titles", "🔍 Why These Recommendations"])
-                                
-                                with sim_tab1:
-                                    for sim_content in similar_content:
-                                        with st.container():
-                                            col1, col2 = st.columns([3, 1])
-                                            with col1:
-                                                st.write(f"**{sim_content.get('title')}** ({sim_content.get('type')}, {sim_content.get('release_year')})")
-                                                st.write(f"*Available on:* {sim_content.get('streaming_service')}")
-                                                st.write(f"*Genres:* {sim_content.get('listed_in')}")
-                                            with col2:
-                                                st.write(f"**Match:** {sim_content.get('similarity', 'N/A')}")
-                                            
-                                            with st.expander("See more details"):
-                                                st.write(f"**Cast:** {sim_content.get('cast', 'N/A')}")
-                                                st.write(f"**Director:** {sim_content.get('director', 'N/A')}")
-                                                st.write(f"**Description:** {sim_content.get('description', 'N/A')}")
-                                            st.write("---")
-                                
-                                with sim_tab2:
-                                    st.write("**Why these recommendations?**")
-                                    st.write("""
-                                    These recommendations are pre-computed using advanced content analysis:
-                                    - Genre and theme matching
-                                    - Plot similarity analysis
-                                    - Cast and director relationships
-                                    - Release year proximity
-                                    
-                                    The percentage match indicates how similar each title is to your selection.
-                                    """)
-                            else:
-                                st.info("No similar content found.")
-            
-            # Bottom pagination controls with better layout
-            st.write("---")
-            page_cols = st.columns([1, 2, 1, 2, 1])
-            
-            # Previous button
-            with page_cols[0]:
-                if st.button("← Previous", disabled=st.session_state.current_page == 1, use_container_width=True):
-                    st.session_state.current_page -= 1
-                    st.rerun()
-            
-            # Spacer
-            with page_cols[1]:
-                st.write("")
-            
-            # Page input
-            with page_cols[2]:
-                page_input = st.number_input(
-                    f"Page (of {total_pages})",
-                    min_value=1,
-                    max_value=total_pages,
-                    value=st.session_state.current_page,
-                    key="page_number",
-                    help=f"Enter a page number between 1 and {total_pages}"
-                )
-                if page_input != st.session_state.current_page:
-                    st.session_state.current_page = page_input
-                    st.rerun()
-            
-            # Spacer
-            with page_cols[3]:
-                st.write("")
-            
-            # Next button
-            with page_cols[4]:
-                if st.button("Next →", disabled=st.session_state.current_page == total_pages, use_container_width=True):
-                    st.session_state.current_page += 1
-                    st.rerun()
-        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 streaming 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(st.session_state.filtered_content)
-                    
-                    try:
-                        # Get AI response
-                        ai_response = get_ai_response(client, user_question, context, model_choice)
-                        
-                        st.success("**AI Assistant Response:**")
-                        st.write(ai_response)
-                        
-                        # Show debug info
-                        with st.expander("Debug Info"):
-                            st.write(f"Model used: {model_choice}")
-                            st.write(f"Content items analyzed: {len(st.session_state.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 st.session_state.filtered_content:
-                            movies = sum(1 for c in st.session_state.filtered_content if c.get('type') == 'Movie')
-                            shows = sum(1 for c in st.session_state.filtered_content if c.get('type') == 'TV Show')
-                            st.write(f"• Found {len(st.session_state.filtered_content)} titles ({movies} movies, {shows} TV shows)")
-                            
-                            # Show top genres
-                            genres = [g.strip() for c in st.session_state.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 st.session_state.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 with improved layout
-    st.markdown("---")
-    if all_content:
-        total_items = len(all_content)
-        filtered_items = len(st.session_state.filtered_content)
-        
-        # Create columns for stats with better spacing
-        stat_cols = st.columns(len(selected_services) + 3)
-        
-        # Basic stats with improved formatting
-        with stat_cols[0]:
-            st.metric("📊 Total Available", f"{total_items:,}")
-        
-        with stat_cols[1]:
-            st.metric("🔍 Filtered Results", f"{filtered_items:,}")
-        
-        with stat_cols[2]:
-            movies = sum(1 for c in st.session_state.filtered_content if c.get('type') == 'Movie')
-            shows = sum(1 for c in st.session_state.filtered_content if c.get('type') == 'TV Show')
-            st.metric("🎬 Movies / 📺 Shows", f"{movies:,} / {shows:,}")
-        
-        # Streaming service breakdown with icons
-        service_icons = {
-            "Netflix": "🔴",
-            "Amazon Prime": "🔵",
-            "Hulu": "🟢",
-            "Disney+": "🟣"
-        }
-        
-        for i, service in enumerate(selected_services, 3):
-            if i < len(stat_cols):
-                service_count = sum(1 for c in st.session_state.filtered_content if c.get('streaming_service') == service)
-                icon = service_icons.get(service, "📺")
-                with stat_cols[i]:
-                    st.metric(f"{icon} {service}", f"{service_count:,}")
-
-if __name__ == "__main__":
-    main() 
+🔍 Searching for similar content to: Dick Johnson Is Dead (ID: s1)
+
+📋 Table structure for mp7bnn9tzhojh7k:
+
+Columns:
+
+[
+0:"Id"
+1:"CreatedAt"
+2:"UpdatedAt"
+3:"title"
+4:"show_id"
+5:"similar_items"
+]
+Sample row:
+
+{
+"Id":1
+"CreatedAt":"2025-06-02 09:34:10+00:00"
+"UpdatedAt":NULL
+"title":"Dick Johnson Is Dead"
+"show_id":"s1"
+"similar_items":"[{"title": "Triple Threat", "show_id": "s3718", "similarity": 0.7407444080262803}, {"title": "The Death and Life of Marsha P. Johnson", "show_id": "s5234", "similarity": 0.7343378197080184}, {"title": "Toxic Beauty", "show_id": "s1715", "similarity": 0.7338750075490208}, {"title": "The Possession of Michael King", "show_id": "s8177", "similarity": 0.731497980641655}, {"title": "Suspiria", "show_id": "s4051", "similarity": 0.7215465108599116}, {"title": "The Sheik", "show_id": "s3726", "similarity": 0.7167594371495939}, {"title": "Operation Avalanche", "show_id": "s8106", "similarity": 0.7060483371270012}, {"title": "Trying Grace", "show_id": "s7595", "similarity": 0.7044626409776164}, {"title": "Shanghai Disney Resort Grand Opening Gala", "show_id": "s295", "similarity": 0.7035801295960805}, {"title": "The Legend of Swee' Pea", "show_id": "s565", "similarity": 0.6992762418494328}, {"title": "The Z Virus", "show_id": "s8732", "similarity": 0.6908215252006857}, {"title": "Extremis", "show_id": "s5798", "similarity": 0.6867877195283146}, {"title": "Journey to Royal: A WWII Rescue Mission", "show_id": "s4807", "similarity": 0.6830348539927185}, {"title": "Narbachi Wadi", "show_id": "s5879", "similarity": 0.6824890235942042}, {"title": "Daredevil DIRECTOR'S CUT", "show_id": "s8325", "similarity": 0.6810832334129707}, {"title": "Aurinko In Adagio: A Rising Voices Film", "show_id": "s5974", "similarity": 0.6756892776605048}, {"title": "Mad Money", "show_id": "s7370", "similarity": 0.6744356204785764}, {"title": "Anjelah Johnson: Not Fancy", "show_id": "s5895", "similarity": 0.6715562800569002}, {"title": "BELLATOR MMA: Kongo vs. Johnson 2", "show_id": "s38", "similarity": 0.6715470446362188}, {"title": "Knives Out", "show_id": "s5656", "similarity": 0.6690988586261262}]"
+}
+📡 Querying table: mp7bnn9tzhojh7k
+
+🔍 Query params: {'where': '(show_id,eq,"s1")'}
+
+📥 Response status: 200
+
+📥 Request URL: https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/mp7bnn9tzhojh7k?where=%28show_id%2Ceq%2C%22s1%22%29
+
+🔄 Trying alternative query format...
+
+📥 Alt response status: 200
+
+📥 Alt request URL: https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/mp7bnn9tzhojh7k?where=%28show_id%2Clike%2C%22s1%22%29
+
+Raw response:
+
+{
+"list":[]
+"pageInfo":{
+"totalRows":0
+"page":1
+"pageSize":25
+"isFirstPage":true
+"isLastPage":true
+}
+}
+❌ No entries found in response
+
+Could not parse response as JSON: {"list":[],"pageInfo":{"totalRows":0,"page":1,"pageSize":25,"isFirstPage":true,"isLastPage":true}}
+
+📊 Found 0 matches in table
+
+📋 Table structure for m8e5rglns4acmef:
+
+Columns:
+
+[
+0:"Id"
+1:"CreatedAt"
+2:"UpdatedAt"
+3:"title"
+4:"show_id"
+5:"similar_items"
+]
+Sample row:
+
+{
+"Id":1
+"CreatedAt":"2025-06-02 09:34:10+00:00"
+"UpdatedAt":NULL
+"title":"Man Vs."
+"show_id":"s1751"
+"similar_items":"[{"title": "Settlers", "show_id": "s5", "similarity": 0.7325137795194292}, {"title": "Nature's Weirdest Events", "show_id": "s7575", "similarity": 0.7188299496186253}, {"title": "After Darkness", "show_id": "s2136", "similarity": 0.7141004186676579}, {"title": "Somnus", "show_id": "s6048", "similarity": 0.7099398640829218}, {"title": "Little John", "show_id": "s6801", "similarity": 0.7046743432606751}, {"title": "Shanghai Disney Resort Grand Opening Gala", "show_id": "s295", "similarity": 0.6926804611909125}, {"title": "Before I'm Dead", "show_id": "s5522", "similarity": 0.6857059160224453}, {"title": "Hostile", "show_id": "s4064", "similarity": 0.6842681482307864}, {"title": "The Haunted Hotel", "show_id": "s6793", "similarity": 0.6802541467049906}, {"title": "Star Wars: The Rise of Skywalker (Episode IX)", "show_id": "s531", "similarity": 0.6760391454742809}, {"title": "High-Rise Invasion", "show_id": "s1045", "similarity": 0.6680968736571052}, {"title": "Humans", "show_id": "s2168", "similarity": 0.6671485424786042}, {"title": "The Handmaid's Tale", "show_id": "s764", "similarity": 0.666126348250307}, {"title": "Adam Ruins Everything", "show_id": "s6090", "similarity": 0.6625023465211879}, {"title": "Coherence", "show_id": "s2812", "similarity": 0.6624407207293317}, {"title": "The Vast of Night", "show_id": "s403", "similarity": 0.6600187026296738}, {"title": "Five Grand", "show_id": "s6647", "similarity": 0.6577227564879091}, {"title": "Technotise: Edit & I", "show_id": "s4188", "similarity": 0.6560547168812791}, {"title": "Ghost Walk", "show_id": "s6779", "similarity": 0.6543972382095791}, {"title": "Beyond the Sky", "show_id": "s3086", "similarity": 0.6536638553019316}]"
+}
+📡 Querying table: m8e5rglns4acmef
+
+🔍 Query params: {'where': '(show_id,eq,"s1")'}
+
+📥 Response status: 200
+
+📥 Request URL: https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/m8e5rglns4acmef?where=%28show_id%2Ceq%2C%22s1%22%29
+
+🔄 Trying alternative query format...
+
+📥 Alt response status: 200
+
+📥 Alt request URL: https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/m8e5rglns4acmef?where=%28show_id%2Clike%2C%22s1%22%29
+
+Raw response:
+
+{
+"list":[]
+"pageInfo":{
+"totalRows":0
+"page":1
+"pageSize":25
+"isFirstPage":true
+"isLastPage":true
+}
+}
+❌ No entries found in response
+
+Could not parse response as JSON: {"list":[],"pageInfo":{"totalRows":0,"page":1,"pageSize":25,"isFirstPage":true,"isLastPage":true}}
+
+📊 Found 0 matches in table
+
+📋 Table structure for m2driodimid10k6:
+
+Columns:
+
+[
+0:"Id"
+1:"CreatedAt"
+2:"UpdatedAt"
+3:"title"
+4:"show_id"
+5:"similar_items"
+]
+Sample row:
+
+{
+"Id":1
+"CreatedAt":"2025-06-02 09:34:09+00:00"
+"UpdatedAt":NULL
+"title":"Make This Tonight"
+"show_id":"s1281"
+"similar_items":"[{"title": "Cake Wars", "show_id": "s2428", "similarity": 0.8643776874920789}, {"title": "Tasty 101", "show_id": "s2270", "similarity": 0.851035239717973}, {"title": "Halloween Wars", "show_id": "s1356", "similarity": 0.839582130187393}, {"title": "Giada's Holiday Handbook", "show_id": "s1251", "similarity": 0.8264038541160613}, {"title": "On Chesil Beach", "show_id": "s212", "similarity": 0.821701576024188}, {"title": "Raw. Vegan. Not Gross.", "show_id": "s2018", "similarity": 0.8133997169642753}, {"title": "Crime Scene Kitchen", "show_id": "s684", "similarity": 0.8130779826834713}, {"title": "Worst Cooks in America", "show_id": "s592", "similarity": 0.8117540162729204}, {"title": "Cutthroat Kitchen", "show_id": "s587", "similarity": 0.8111139381056529}, {"title": "Lightened Up", "show_id": "s1280", "similarity": 0.8079789805365294}, {"title": "Chuck's World", "show_id": "s216", "similarity": 0.8049357265970711}, {"title": "Ayesha's Home Kitchen", "show_id": "s1240", "similarity": 0.8013048453237337}, {"title": "Struggle Meals", "show_id": "s1283", "similarity": 0.793547594830839}, {"title": "TrueSouth", "show_id": "s869", "similarity": 0.7906473317607992}, {"title": "Behind The Dish", "show_id": "s1279", "similarity": 0.7876233721299817}, {"title": "Eater's Guide to the World", "show_id": "s1212", "similarity": 0.7800859117925178}, {"title": "Supermarket Stakeout", "show_id": "s1367", "similarity": 0.7786342300343229}, {"title": "The Grill Iron", "show_id": "s1284", "similarity": 0.7781488386934403}, {"title": "F*ck, That's Delicious", "show_id": "s1308", "similarity": 0.7778297208530904}, {"title": "Jamie: Keep Cooking and Carry On", "show_id": "s1496", "similarity": 0.7756508905864695}]"
+}
+📡 Querying table: m2driodimid10k6
+
+🔍 Query params: {'where': '(show_id,eq,"s1")'}
+
+📥 Response status: 200
+
+📥 Request URL: https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/m2driodimid10k6?where=%28show_id%2Ceq%2C%22s1%22%29
+
+🔄 Trying alternative query format...
+
+📥 Alt response status: 200
+
+📥 Alt request URL: https://mtoft20-potm.hf.space/api/v1/db/data/noco/p9pozkcw81t9aee/m2driodimid10k6?where=%28show_id%2Clike%2C%22s1%22%29
+
+Raw response:
+
+{
+"list":[]
+"pageInfo":{
+"totalRows":0
+"page":1
+"pageSize":25
+"isFirstPage":true
+"isLastPage":true
+}
+}
+❌ No entries found in response
+
+Could not parse response as JSON: {"list":[],"pageInfo":{"totalRows":0,"page":1,"pageSize":25,"isFirstPage":true,"isLastPage":true}}
+
+📊 Found 0 matches in table
+
+📝 Total similar items found: 0
+
+No similar content found.