import openai
# NEW: from openai import OpenAI
from llama_index import (
    SimpleDirectoryReader, GPTVectorStoreIndex, LLMPredictor, PromptHelper,
    VectorStoreIndex, load_index_from_storage, StorageContext, Prompt
)
from llama_index.prompts import ChatPromptTemplate, ChatMessage, MessageRole
from langchain.chat_models import ChatOpenAI
import gradio as gr
import logging
import os

# OpenAI client initialization
# client = OpenAI(api_key=os.environ['OPENAI_API_KEY'])

api_key=os.environ['OPENAI_API_KEY']

# Set up logging configuration
logging.basicConfig(filename='Questions_asked.log', level=logging.INFO, format='%(asctime)s - %(message)s', datefmt='%Y-%m-%d %H:%M:%S')

def construct_index(directory_path):
    max_input_size = 4896
    num_outputs = 512
    max_chunk_overlap = 20
    chunk_size_limit = 681
    prompt_helper = PromptHelper(max_input_size, num_outputs, chunk_overlap_ratio=0.1, chunk_size_limit=chunk_size_limit)

    llm_predictor = LLMPredictor(llm=ChatOpenAI(temperature=0.7, model_name="gpt-3.5-turbo", max_tokens=num_outputs))

    documents = SimpleDirectoryReader(directory_path).load_data()

    # INDEX
    index = GPTVectorStoreIndex.from_documents(documents, llm_predictor=llm_predictor, prompt_helper=prompt_helper)

    # SAVE
    index.set_index_id("vector_index")
    index.storage_context.persist("./storage")

    return index

def chatbot(input_text):
    # rebuild storage context
    storage_context = StorageContext.from_defaults(persist_dir="storage")
    # load index
    index = load_index_from_storage(storage_context, index_id="vector_index")

    # Create response template
    TEMPLATE_STR=(
    "I am a bot specialized in answering questions related to SATS. Please ask your question in any language, and I will respond in the same language. For more precise answers, include specific details in your query.\n"
    "---------------------\n"
    "{context_str}"
    "\n---------------------\n"
    "Given the context provided above, I will now address your question: {query_str}. For yes/no questions, I will offer further explanation or guidance. If the answer is beyond my capabilities, I will inform you."
    )

    QA_TEMPLATE = Prompt(TEMPLATE_STR)
    query_engine = index.as_query_engine(text_qa_template=QA_TEMPLATE, response_mode="compact")

    response = query_engine.query(input_text)

    return response.response

# Construct the index with documents directory
index = construct_index("docs")

# Gradio interface
iface = gr.Interface(
    fn=chatbot,
    inputs=gr.components.Textbox(lines=7, label="Ask with a spesific case or about rules on a subject"),
    outputs="text",
    title="SATS AI"
)
iface.launch(share=False, auth=(os.environ.get("User"), os.environ.get("Password")))