app.py

import streamlit as st
import os
import tempfile
from dotenv import load_dotenv

# Load environment variables
load_dotenv()

# Import required libraries
try:
    from langchain_community.document_loaders import PyPDFLoader
    from langchain_text_splitters import RecursiveCharacterTextSplitter
    from langchain_core.documents import Document
    from langchain_community.retrievers import BM25Retriever
    from smolagents import Tool, CodeAgent, InferenceClientModel
except ImportError as e:
    st.error(f"Missing dependency: {e}. Please install all requirements.")
    st.stop()

# Custom Retriever Tool
class RetrieverTool(Tool):
    name = "retriever"
    description = "Uses semantic search to retrieve the parts of the research paper that could be most relevant to answer your query."
    inputs = {
        "query": {
            "type": "string",
            "description": "The query to perform. This should be semantically close to your target documents. Use the affirmative form rather than a question.",
        }
    }
    output_type = "string"

    def __init__(self, docs, **kwargs):
        super().__init__(**kwargs)
        self.retriever = BM25Retriever.from_documents(docs, k=10)

    def forward(self, query: str) -> str:
        """Execute the retrieval based on the provided query."""
        assert isinstance(query, str), "Your search query must be a string"
        
        docs = self.retriever.invoke(query)
        
        return "\nRetrieved documents:\n" + "".join([
            f"\n\n===== Document {str(i)} (Page {doc.metadata.get('page', 'N/A')}) =====\n" + doc.page_content
            for i, doc in enumerate(docs)
        ])

# Function to load and process PDF
@st.cache_resource
def load_and_process_pdf(pdf_path):
    """Load PDF and split into chunks for retrieval."""
    try:
        # Load PDF
        loader = PyPDFLoader(pdf_path)
        pages = loader.load()
        
        # Split into chunks
        text_splitter = RecursiveCharacterTextSplitter(
            chunk_size=500,
            chunk_overlap=50,
            add_start_index=True,
            strip_whitespace=True,
            separators=["\n\n", "\n", ".", " ", ""],
        )
        docs_processed = text_splitter.split_documents(pages)
        
        return docs_processed, len(pages)
    except Exception as e:
        st.error(f"Error processing PDF: {e}")
        return None, 0

# Function to create agent
@st.cache_resource
def create_agent(_docs):
    """Create the RAG agent with retriever tool."""
    retriever_tool = RetrieverTool(_docs)
    
    # Use FREE Hugging Face model (Qwen 2.5 72B via serverless inference)
    agent = CodeAgent(
        tools=[retriever_tool],
        model=InferenceClientModel(
            model_id="Qwen/Qwen2.5-72B-Instruct",
            token=os.getenv("HF_TOKEN")
        ),
        max_steps=4,
        verbosity_level=0,
    )
    
    return agent

# Streamlit UI
def main():
    st.set_page_config(
        page_title="PaperChat",
        page_icon="📄",
        layout="wide"
    )
    
    # Header
    st.title("📄 PaperChat - Research Paper Q&A Assistant")
    st.markdown("""
    Upload any research paper (PDF) and ask questions about it. 
    Powered by Agentic RAG with retrieval capabilities.
    """)
    
    # Sidebar
    with st.sidebar:
        st.header("📤 Upload Paper")
        uploaded_file = st.file_uploader(
            "Choose a PDF file",
            type="pdf",
            help="Upload a research paper in PDF format"
        )
        
        st.markdown("---")
        st.subheader("📚 Example Questions")
        st.markdown("""
        - What is the main contribution of this paper?
        - What methodology was used?
        - What are the key results?
        - What datasets were used?
        - What are the limitations mentioned?
        """)
        
        st.markdown("---")
        st.subheader("ℹ️ How it works")
        st.markdown("""
        1. Upload your paper
        2. The system chunks and indexes it
        3. Ask questions naturally
        4. Get answers with source citations
        """)
    
    # Main content area
    if uploaded_file is not None:
        # Save uploaded file to temporary location
        with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file:
            tmp_file.write(uploaded_file.read())
            tmp_path = tmp_file.name
        
        # Process PDF
        with st.spinner("🔄 Processing your paper... This may take a moment."):
            docs, num_pages = load_and_process_pdf(tmp_path)
        
        if docs:
            st.success(f"✅ Paper loaded successfully! ({num_pages} pages, {len(docs)} chunks)")
            
            # Create agent
            with st.spinner("🤖 Initializing AI agent..."):
                agent = create_agent(docs)
            
            st.success("✅ Agent ready! You can now ask questions.")
            
            # Chat interface
            st.markdown("---")
            st.subheader("💬 Ask Questions")
            
            # Initialize chat history
            if "messages" not in st.session_state:
                st.session_state.messages = []
            
            # Display chat history
            for message in st.session_state.messages:
                with st.chat_message(message["role"]):
                    st.markdown(message["content"])
            
            # Chat input
            if question := st.chat_input("Ask a question about the paper..."):
                # Add user message to chat history
                st.session_state.messages.append({"role": "user", "content": question})
                
                # Display user message
                with st.chat_message("user"):
                    st.markdown(question)
                
                # Generate response
                with st.chat_message("assistant"):
                    with st.spinner("🤔 Thinking..."):
                        try:
                            answer = agent.run(question)
                            st.markdown(answer)
                            
                            # Add assistant response to chat history
                            st.session_state.messages.append({"role": "assistant", "content": answer})
                        except Exception as e:
                            error_msg = f"Error generating answer: {str(e)}"
                            st.error(error_msg)
                            st.session_state.messages.append({"role": "assistant", "content": error_msg})
            
            # Clear chat button
            if st.button("🗑️ Clear Chat History"):
                st.session_state.messages = []
                st.rerun()
        
        # Cleanup temp file
        try:
            os.unlink(tmp_path)
        except:
            pass
    
    else:
        # Welcome message when no file is uploaded
        st.info("👈 Please upload a research paper PDF from the sidebar to get started.")
        
        st.markdown("### 🎯 What can you do with PaperChat?")
        col1, col2, col3 = st.columns(3)
        
        with col1:
            st.markdown("""
            #### 📖 Understand Papers
            - Get summaries of complex papers
            - Understand methodology
            - Learn about key findings
            """)
        
        with col2:
            st.markdown("""
            #### 🔍 Extract Information
            - Find specific details
            - Locate datasets used
            - Identify citations
            """)
        
        with col3:
            st.markdown("""
            #### 💡 Learn Faster
            - Ask follow-up questions
            - Clarify concepts
            - Compare approaches
            """)

if __name__ == "__main__":
    main()