app.py

#from utils.credentials import check_credentials, init_clients

import os
import streamlit as st
from langchain.chains import RetrievalQA
from PyPDF2 import PdfReader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.callbacks.base import BaseCallbackHandler
from langchain.vectorstores.neo4j_vector import Neo4jVector
from streamlit.logger import get_logger
from chains import (
    load_embedding_model,
    load_llm,
)
from pymongo import MongoClient
import certifi

#url = os.getenv("NEO4J_URI")
#username = os.getenv("NEO4J_USERNAME")
#password = os.getenv("NEO4J_PASSWORD")

#url = os.getenv("MONGO_URI")
#username = os.getenv("NEO4J_USERNAME")
#password = os.getenv("NEO4J_PASSWORD")

import os
from pymongo import MongoClient
from openai import OpenAI
#from dotenv import load_dotenv

# Load environment variables
#load_dotenv()

# Initialize clients
#openai_client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
#atlas_uri = os.getenv("ATLAS_URI")
#url = atlas_uri
#client = MongoClient(atlas_uri)


import requests
from pymongo import MongoClient
import certifi

# Connect to MongoDB Atlas
#client = MongoClient(atlas_uri,tls=True,tlsCAFile=certifi.where())
#db = client['sample_mflix']
#collection = db['embedded_movies']


ollama_base_url = os.getenv("OLLAMA_BASE_URL")
embedding_model_name = os.getenv("EMBEDDING_MODEL", "SentenceTransformer" )
llm_name = os.getenv("LLM", "llama2")
#url = os.getenv("NEO4J_URI")

# Check if the required environment variables are set
#if not all([url, username, password,
 #         ollama_base_url]):
if not all([
          ollama_base_url]):
    st.write("The application requires some information before running.")
    with st.form("connection_form"):
        #url = st.text_input("Enter ATLAS_URI",)
        #username = st.text_input("Enter NEO4J_USERNAME")
        #password = st.text_input("Enter NEO4J_PASSWORD", type="password")
        ollama_base_url = st.text_input("Enter OLLAMA_BASE_URL")
        st.markdown("Only enter the OPENAI_APIKEY to use OpenAI instead of Ollama. Leave blank to use Ollama.")
        openai_apikey = st.text_input("Enter OPENAI_API_KEY", type="password")
        submit_button = st.form_submit_button("Submit")
    if submit_button:
        #if not all([url, username, password, ]):
        #if not all([url, ]):
        #    st.write("Enter the ATLAS information.")
        if not (ollama_base_url or openai_apikey):
            st.write("Enter the Ollama URL or OpenAI API Key.")
        if openai_apikey:
            llm_name = "gpt-3.5"
            os.environ['OPENAI_API_KEY'] = openai_apikey

#os.environ["NEO4J_URL"] = url
#os.environ["ATLAS_URI"] = url

logger = get_logger(__name__)

embeddings, dimension = load_embedding_model(
    embedding_model_name, config={"ollama_base_url": ollama_base_url}, logger=logger
)


class StreamHandler(BaseCallbackHandler):
    def __init__(self, container, initial_text=""):
        self.container = container
        self.text = initial_text

    def on_llm_new_token(self, token: str, **kwargs) -> None:
        self.text += token
        self.container.markdown(self.text)

llm = load_llm(llm_name, logger=logger, config={"ollama_base_url": ollama_base_url})


def main():
        st.header("📄Chat with your pdf file")

        # upload a your pdf file
        pdf = st.file_uploader("Upload your PDF", type="pdf")

        if pdf is not None:
            pdf_reader = PdfReader(pdf)

            text = ""
            for page in pdf_reader.pages:
                text += page.extract_text()

            # langchain_textspliter
            text_splitter = RecursiveCharacterTextSplitter(
                chunk_size=1000, chunk_overlap=200, length_function=len
            )

            chunks = text_splitter.split_text(text=text)

            # Store the chunks part in db (vector)
            vectorstore = Neo4jVector.from_texts(
                chunks,
                url=url,
                username=username,
                password=password,
                embedding=embeddings,
                index_name="pdf_bot",
                node_label="PdfBotChunk",
                pre_delete_collection=True,  # Delete existing PDF data
            )
            qa = RetrievalQA.from_chain_type(
                llm=llm, chain_type="stuff", retriever=vectorstore.as_retriever()
            )

            # Accept user questions/query
            query = st.text_input("Ask questions about your PDF file")

            if query:
                stream_handler = StreamHandler(st.empty())
                qa.run(query, callbacks=[stream_handler])


if __name__ == "__main__":
     main()