app.py

import streamlit as st
from PyPDF2 import PdfReader
from io import BytesIO
import os
import tempfile

# Fixed imports for LangChain
from langchain_text_splitters import RecursiveCharacterTextSplitter
from langchain_google_genai import GoogleGenerativeAIEmbeddings
from langchain_community.vectorstores import FAISS
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain.chains.question_answering import load_qa_chain
from langchain.prompts import PromptTemplate

# --- Get API key from Hugging Face Secrets ---
# In Hugging Face Spaces, set this in Settings -> Repository secrets
GOOGLE_API_KEY = os.environ.get("GOOGLE_API_KEY", "")

# Use temporary directory for Hugging Face Spaces
TEMP_DIR = tempfile.gettempdir()
FAISS_INDEX_PATH = os.path.join(TEMP_DIR, "faiss_index")

def get_pdf_text(pdf_docs):
    text = ""
    for pdf in pdf_docs:
        pdf_reader = PdfReader(BytesIO(pdf.read()))
        for page in pdf_reader.pages:
            page_text = page.extract_text()
            if page_text:
                text += page_text
    return text

def get_text_chunks(text):
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=10000, chunk_overlap=1000)
    return text_splitter.split_text(text)

def get_vector_store(text_chunks, api_key):
    embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001", google_api_key=api_key)
    vector_store = FAISS.from_texts(text_chunks, embedding=embeddings)
    vector_store.save_local(FAISS_INDEX_PATH)

def get_conversational_chain(api_key):
    prompt_template = """
    You are a helpful assistant that only answers based on the context provided from the PDF documents.
    Do not use any external knowledge or assumptions. If the answer is not found in the context below, reply with "I don't know."
    Context:
    {context}
    Question:
    {question}
    Answer:
    """
    model = ChatGoogleGenerativeAI(model="gemini-2.0-flash-exp", temperature=0, google_api_key=api_key)
    prompt = PromptTemplate(template=prompt_template, input_variables=["context", "question"])
    chain = load_qa_chain(model, chain_type="stuff", prompt=prompt)
    return chain

def user_input(user_question, api_key):
    embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001", google_api_key=api_key)
    new_db = FAISS.load_local(FAISS_INDEX_PATH, embeddings, allow_dangerous_deserialization=True)
    docs = new_db.similarity_search(user_question)
    chain = get_conversational_chain(api_key)
    response = chain({"input_documents": docs, "question": user_question}, return_only_outputs=True)
    st.write("Reply: ", response["output_text"])

def main():
    st.set_page_config(page_title="Chat PDF", page_icon="")
    st.header("Retrieval-Augmented Generation - Gemini 2.0")
    st.markdown("---")

    # Initialize session state
    if "api_entered" not in st.session_state:
        st.session_state["api_entered"] = False
    if "pdf_processed" not in st.session_state:
        st.session_state["pdf_processed"] = False

    # Check for API key
    api_key = GOOGLE_API_KEY

    # STEP 1: API Key handling
    if not st.session_state["api_entered"]:
        if not api_key:
            st.warning(" Google API Key not found in environment variables.")
            st.info("Please add GOOGLE_API_KEY to your Hugging Face Space secrets or enter it below.")
            user_api_key = st.text_input("Enter your Gemini API key", type="password", help="Get your API key from https://makersuite.google.com/app/apikey")
            if st.button("Continue", type="primary") and user_api_key:
                st.session_state["user_api_key"] = user_api_key
                st.session_state["api_entered"] = True
                st.rerun()
            st.stop()
        else:
            st.session_state["user_api_key"] = api_key
            st.session_state["api_entered"] = True

    api_key = st.session_state.get("user_api_key", "")

    # STEP 2: Upload PDF(s)
    if not st.session_state["pdf_processed"]:
        st.subheader(" Step 1: Upload your PDF file(s)")
        pdf_docs = st.file_uploader(
            "Upload PDF files", 
            accept_multiple_files=True, 
            type=['pdf'],
            help="Select one or more PDF files to analyze"
        )
        
        if st.button("Submit & Process PDFs", type="primary", disabled=not pdf_docs):
            if pdf_docs:
                with st.spinner("Processing PDFs... This may take a moment."):
                    try:
                        raw_text = get_pdf_text(pdf_docs)
                        if not raw_text.strip():
                            st.error(" No text could be extracted from the PDF(s). Please check your files.")
                            st.stop()
                        
                        text_chunks = get_text_chunks(raw_text)
                        get_vector_store(text_chunks, api_key)
                        st.session_state["pdf_processed"] = True
                        st.success(" PDFs processed successfully! You can now ask questions.")
                        st.rerun()
                    except Exception as e:
                        st.error(f" Error processing PDFs: {str(e)}")
                        st.stop()
            else:
                st.error("Please upload at least one PDF file.")
        
        if not pdf_docs:
            st.info(" Please upload one or more PDF files to get started.")
        
        st.stop()

    # STEP 3: Ask questions
    st.subheader(" Step 2: Ask questions about your PDFs")
    
    # Add a reset button
    col1, col2 = st.columns([3, 1])
    with col2:
        if st.button(" Upload New PDFs"):
            st.session_state["pdf_processed"] = False
            st.rerun()
    
    # Question input
    user_question = st.text_input(
        "Ask a question about your uploaded PDFs",
        placeholder="e.g., What are the main topics discussed in the document?",
        help="The AI will only answer based on the content of your uploaded PDFs"
    )
    
    if user_question:
        with st.spinner("Searching for answer..."):
            try:
                user_input(user_question, api_key)
            except Exception as e:
                st.error(f" Error getting answer: {str(e)}")

    # Add footer
    st.markdown("---")
    st.markdown(
        """
        <div style='text-align: center'>
            <small></small>
        </div>
        """, 
        unsafe_allow_html=True
    )

if __name__ == "__main__":
    main()