app.py

import os
import streamlit as st
import pdfplumber
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import FAISS
from transformers import pipeline, M2M100ForConditionalGeneration, AutoTokenizer

# Set up the page configuration
st.set_page_config(page_title="RAG-based PDF Chat", layout="centered", page_icon="📄")

# Load the summarization pipeline model
@st.cache_resource
def load_summarization_pipeline():
    return pipeline("summarization", model="facebook/bart-large-cnn")

summarizer = load_summarization_pipeline()

# Load the translation model
@st.cache_resource
def load_translation_model():
    model = M2M100ForConditionalGeneration.from_pretrained("alirezamsh/small100")
    tokenizer = AutoTokenizer.from_pretrained("alirezamsh/small100")
    return model, tokenizer

translation_model, translation_tokenizer = load_translation_model()

# Define available languages for translation
LANGUAGES = {
    "English": "en",
    "French": "fr",
    "Spanish": "es",
    "Chinese": "zh",
    "Hindi": "hi",
    "Urdu": "ur",
}

# Split text into manageable chunks
@st.cache_data
def get_text_chunks(text):
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=10000, chunk_overlap=1000)
    return text_splitter.split_text(text)

# Initialize embedding function
embedding_function = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")

# Create a FAISS vector store with embeddings
@st.cache_resource
def load_or_create_vector_store(text_chunks):
    return FAISS.from_texts(text_chunks, embedding=embedding_function) if text_chunks else None

# Helper function to process a single PDF
def process_single_pdf(file_path):
    text = ""
    try:
        with pdfplumber.open(file_path) as pdf:
            for page in pdf.pages:
                page_text = page.extract_text()
                if page_text:
                    text += page_text
    except Exception as e:
        st.error(f"Failed to read PDF: {file_path} - {e}")
    return text

# Load PDFs with progress display
def load_pdfs_with_progress(folder_path):
    if not os.path.exists(folder_path):
        st.error(f"The folder '{folder_path}' does not exist. Please create it and add PDF files.")
        return None

    all_text = ""
    pdf_files = [os.path.join(folder_path, filename) for filename in os.listdir(folder_path) if filename.endswith('.pdf')]
    if not pdf_files:
        st.error("No PDF files found in the specified folder.")
        return None

    st.markdown("### Loading data...")
    progress_bar = st.progress(0)

    for i, file_path in enumerate(pdf_files):
        all_text += process_single_pdf(file_path)
        progress_bar.progress((i + 1) / len(pdf_files))

    progress_bar.empty()
    return load_or_create_vector_store(get_text_chunks(all_text)) if all_text else None

# Generate summary based on retrieved text
def generate_summary(query, retrieved_text):
    summarization_input = f"{query} Related information:{retrieved_text}"[:1024]
    summary = summarizer(summarization_input, max_length=500, min_length=50, do_sample=False)
    return summary[0]["summary_text"]

# Translate text to selected language
def translate_text(text, target_lang_code):
    # Set the target language
    translation_tokenizer.src_lang = "en"  # assuming the original language is English
    translation_tokenizer.tgt_lang = target_lang_code  # dynamically use the selected target language code

    # Tokenize the text and generate the translation
    encoded_text = translation_tokenizer(text, return_tensors="pt")
    generated_tokens = translation_model.generate(**encoded_text, forced_bos_token_id=translation_tokenizer.lang_code_to_id[target_lang_code])
    return translation_tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)[0]

# Main function to run the Streamlit app
def main():
    st.markdown(
        """
        <h1 style="font-size:30px; text-align: center;">
        📄 JusticeCompass: Your AI-Powered Legal Navigator for Swift, Accurate Guidance.
        </h1>
        """,
        unsafe_allow_html=True
    )

    if "vector_store" not in st.session_state:
        st.session_state["vector_store"] = load_pdfs_with_progress('documents1')
        if st.session_state["vector_store"] is None:
            return

    # Prompt input
    user_question = st.text_input("Ask a Question:", placeholder="Type your question here...")

    # Language selection dropdown
    selected_language = st.selectbox("Select output language:", list(LANGUAGES.keys()))

    if user_question and st.button("Get Response"):
        with st.spinner("Generating response..."):
            docs = st.session_state["vector_store"].similarity_search(user_question)
            context_text = " ".join([doc.page_content for doc in docs])
            answer = generate_summary(user_question, context_text)
            translated_answer = translate_text(answer, LANGUAGES[selected_language])
            st.markdown(f"**🤖 AI ({selected_language}):** {translated_answer}")

if __name__ == "__main__":
    main()