app.py

import streamlit as st
from pypdf import PdfReader
from langchain.text_splitter import CharacterTextSplitter
from langchain.embeddings import OpenAIEmbeddings
from langchain.vectorstores import FAISS
from langchain.chat_models import ChatOpenAI
from langchain.memory import ConversationBufferMemory
from langchain.chains import ConversationalRetrievalChain
from htmlTemplates import css, bot_template, user_template

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

def get_text_chunks(text):
    text_splitter = CharacterTextSplitter(
        separator="\n",
        chunk_size=1000,
        chunk_overlap=200,
        length_function=len
    )
    chunks = text_splitter.split_text(text)
    return chunks

def get_vectorstore(text_chunks, openai_api_key, embedding_model):
    embeddings = OpenAIEmbeddings(api_key=openai_api_key, model=embedding_model)
    vectorstore = FAISS.from_texts(texts=text_chunks, embedding=embeddings)
    return vectorstore

def get_conversation_chain(vectorstore, openai_api_key, chat_model):
    llm = ChatOpenAI(api_key=openai_api_key, model=chat_model)  
    memory = ConversationBufferMemory(
        memory_key='chat_history', return_messages=True)
    conversation_chain = ConversationalRetrievalChain.from_llm(
        llm=llm,
        retriever=vectorstore.as_retriever(),
        memory=memory
    )
    return conversation_chain

def handle_userinput(user_question):
    # Simpan pertanyaan pengguna ke dalam riwayat chat
    st.session_state.chat_history.append({"role": "user", "content": user_question})
    
    # Dapatkan respons dari AI
    response = st.session_state.conversation({'question': user_question})
    st.session_state.chat_history.append({"role": "bot", "content": response['answer']})

    # Tampilkan semua pesan dalam riwayat chat
    for message in st.session_state.chat_history:
        if message['role'] == 'user':
            st.write(user_template.replace("{{MSG}}", message['content']), unsafe_allow_html=True)
        else:
            st.write(bot_template.replace("{{MSG}}", message['content']), unsafe_allow_html=True)

def main():
    st.set_page_config(page_title="Chat with multiple PDFs", page_icon=":books:")
    st.write(css, unsafe_allow_html=True)

    openai_api_key = st.sidebar.text_input("Enter your OpenAI API Key", type="password")

    # Pilihan model untuk embeddings
    embedding_model_options = [
        "text-embedding-3-large",
        "text-embedding-3-small",
        "text-embedding-ada-002"
    ]
    selected_embedding_model = st.sidebar.selectbox("Select the Embedding Model", embedding_model_options)

    # Pilihan model untuk chat
    chat_model_options = [
        "gpt-4o-mini",
        "gpt-3.5-turbo-0125"
    ]
    selected_chat_model = st.sidebar.selectbox("Select the Chat Model", chat_model_options)

    if "conversation" not in st.session_state:
        st.session_state.conversation = None
    if "chat_history" not in st.session_state:
        st.session_state.chat_history = []

    st.header("Chat with multiple PDFs :books:")
    st.write("Please enter the data in the menu on the left")  # Menambahkan teks di sini
    
    # Menggunakan text_area untuk input pengguna
    user_question = st.text_area("Ask a question about your documents:", height=100)

    # Menambahkan tombol untuk mengirim pertanyaan
    if st.button("Send") and user_question and st.session_state.conversation:
        handle_userinput(user_question)
        st.session_state.user_question = ""  # Mengosongkan input setelah mengirim

    with st.sidebar:
        st.subheader("Your documents")
        pdf_docs = st.file_uploader("Upload your PDFs here", accept_multiple_files=True)
        
        if pdf_docs and openai_api_key:
            if st.button("Process PDFs"):
                with st.spinner("Processing"):
                    # get pdf text
                    raw_text = get_pdf_text(pdf_docs)

                    # get the text chunks
                    text_chunks = get_text_chunks(raw_text)

                    # create vector store
                    vectorstore = get_vectorstore(text_chunks, openai_api_key, selected_embedding_model)

                    # create conversation chain
                    st.session_state.conversation = get_conversation_chain(vectorstore, openai_api_key, selected_chat_model)
                    st.success("PDFs processed successfully!")

if __name__ == '__main__':
    main()