app.py

import streamlit as st
import sys
import os
from pathlib import Path

# Add src to path
sys.path.append(str(Path(__file__).parent))

from rag_pipeline import query_rag
from ingest import create_vector_db

# Page configuration
st.set_page_config(
    page_title="RAG Question Answering System",
    page_icon="🤖",
    layout="wide"
)

# Custom CSS
st.markdown("""
<style>
    .main-header {
        font-size: 2.5rem;
        font-weight: 700;
        margin-bottom: 1rem;
    }
    .sub-header {
        font-size: 1.2rem;
        color: #666;
        margin-bottom: 2rem;
    }
    .source-box {
        background-color: #f0f2f6;
        padding: 1rem;
        border-radius: 0.5rem;
        margin-top: 1rem;
    }
</style>
""", unsafe_allow_html=True)

# Header
st.markdown('<div class="main-header">RAG Question Answering System</div>', unsafe_allow_html=True)
st.markdown('<div class="sub-header">Ask questions about your documents using AI</div>', unsafe_allow_html=True)

# Sidebar
with st.sidebar:
    st.header("📚 Document Management")
    
    # File upload
    uploaded_files = st.file_uploader(
        "Upload documents (TXT or PDF)",
        type=['txt', 'pdf'],
        accept_multiple_files=True
    )
    
    if uploaded_files:
        data_dir = Path("data")
        data_dir.mkdir(exist_ok=True)
        
        for uploaded_file in uploaded_files:
            file_path = data_dir / uploaded_file.name
            with open(file_path, "wb") as f:
                f.write(uploaded_file.getbuffer())
        
        st.success(f"Uploaded {len(uploaded_files)} file(s)")
    
    # Ingest button
    if st.button("🔄 Process Documents", type="primary"):
        with st.spinner("Processing documents..."):
            success = create_vector_db()
            if success:
                st.success("Documents processed successfully!")
                st.rerun()
            else:
                st.error("Error processing documents")
    
    st.divider()
    
    # Settings
    st.header("⚙️ Settings")
    st.caption("Current configuration:")
    st.caption("• Chunk size: 500")
    st.caption("• Chunk overlap: 50")
    st.caption("• Top-K retrieval: 3")
    st.caption("• Model: Llama 3.1-8B")

# Main area
tab1, tab2 = st.tabs(["💬 Ask Questions", "📊 About"])

with tab1:
    # Check if vector store exists
    if not Path("faiss_index").exists():
        st.warning("⚠️ No documents have been processed yet. Please upload and process documents first.")
    else:
        # Question input
        question = st.text_input(
            "Enter your question:",
            placeholder="What is quantum computing?",
            key="question_input"
        )
        
        col1, col2 = st.columns([1, 5])
        with col1:
            ask_button = st.button("🔍 Ask", type="primary")
        
        if ask_button and question:
            with st.spinner("Searching documents and generating answer..."):
                answer, sources = query_rag(question)
                
                # Display answer
                st.markdown("### Answer:")
                st.markdown(answer)
                
                # Display sources
                if sources:
                    st.markdown("### Sources:")
                    for i, doc in enumerate(sources, 1):
                        source_file = doc.metadata.get('source', 'Unknown')
                        source_name = Path(source_file).name
                        with st.expander(f"📄 Source {i}: {source_name}"):
                            st.text(doc.page_content[:500] + "..." if len(doc.page_content) > 500 else doc.page_content)
        
        # Example questions
        st.markdown("---")
        st.markdown("#### Example Questions:")
        example_questions = [
            "What is quantum computing?",
            "How does blockchain work?",
            "What are the main causes of climate change?",
            "What is artificial intelligence?"
        ]
        
        cols = st.columns(2)
        for i, eq in enumerate(example_questions):
            with cols[i % 2]:
                if st.button(eq, key=f"example_{i}"):
                    st.session_state.question_input = eq
                    st.rerun()

with tab2:
    st.markdown("""
    ## About This System
    
    This is a **Retrieval-Augmented Generation (RAG)** system that answers questions based on your documents.
    
    ### How It Works:
    
    1. **Upload Documents**: Add your TXT or PDF files
    2. **Process**: The system splits documents into chunks and creates vector embeddings
    3. **Ask Questions**: The system finds relevant chunks and generates accurate answers
    
    ### Technology Stack:
    
    - **LangChain**: RAG framework
    - **FAISS**: Vector database for similarity search
    - **Groq API**: Fast LLM inference (Llama 3.1-8B)
    - **HuggingFace**: Local embeddings
    - **Streamlit**: Web interface
    
    ### Features:
    
    - ✅ PDF and TXT file support
    - ✅ Intelligent chunking strategies
    - ✅ Source attribution
    - ✅ Fast and accurate responses
    - ✅ Free to use (Groq free tier)
    
    ---
    
    **Built with ❤️ using LangChain and Streamlit**
    """)

# Footer
st.markdown("---")
st.caption("💡 Tip: Upload domain-specific documents for more accurate answers in that field.")