import os

import gradio as gr
from text_generation import Client
from conversation import get_conv_template
from transformers import AutoTokenizer
from pymongo import MongoClient

DB_NAME = os.getenv("MONGO_DBNAME", "taiwan-llm")
USER = os.getenv("MONGO_USER")
PASSWORD = os.getenv("MONGO_PASSWORD")

uri = f"mongodb+srv://{USER}:{PASSWORD}@{DB_NAME}.kvwjiok.mongodb.net/?retryWrites=true&w=majority"
mongo_client = MongoClient(uri)
db = mongo_client[DB_NAME]
conversations_collection = db['conversations']

DESCRIPTION = """
# Breeze
"""

LICENSE = """
"""

DEFAULT_SYSTEM_PROMPT = "You are a helpful AI assistant built by MediaTek Research. The user you are helping speaks Traditional Chinese and comes from Taiwan."

endpoint_url = os.environ.get("ENDPOINT_URL", "http://127.0.0.1:8080")
client = Client(endpoint_url, timeout=120)
eos_token = "</s>"
MAX_MAX_NEW_TOKENS = 4096
DEFAULT_MAX_NEW_TOKENS = 1536

max_prompt_length = 8192 - MAX_MAX_NEW_TOKENS - 10

model_name = "yentinglin/Taiwan-LLM-7B-v2.0-chat"
tokenizer = AutoTokenizer.from_pretrained(model_name)

with gr.Blocks() as demo:
    gr.Markdown(DESCRIPTION)

    chatbot = gr.Chatbot()
    with gr.Row():
        msg = gr.Textbox(
            container=False,
            show_label=False,
            placeholder='Type a message...',
            scale=10,
        )
        submit_button = gr.Button('Submit',
                                  variant='primary',
                                  scale=1,
                                  min_width=0)

    with gr.Row():
        retry_button = gr.Button('🔄  Retry', variant='secondary')
        undo_button = gr.Button('↩️ Undo', variant='secondary')
        clear = gr.Button('🗑️  Clear', variant='secondary')

    saved_input = gr.State()

    with gr.Accordion(label='Advanced options', open=False):
        system_prompt = gr.Textbox(label='System prompt',
                                   value=DEFAULT_SYSTEM_PROMPT,
                                   lines=6)
        max_new_tokens = gr.Slider(
            label='Max new tokens',
            minimum=1,
            maximum=MAX_MAX_NEW_TOKENS,
            step=1,
            value=DEFAULT_MAX_NEW_TOKENS,
        )
        temperature = gr.Slider(
            label='Temperature',
            minimum=0.1,
            maximum=1.0,
            step=0.1,
            value=0.3,
        )
        top_p = gr.Slider(
            label='Top-p (nucleus sampling)',
            minimum=0.05,
            maximum=1.0,
            step=0.05,
            value=0.95,
        )
        top_k = gr.Slider(
            label='Top-k',
            minimum=1,
            maximum=1000,
            step=1,
            value=50,
        )

    def user(user_message, history):
        return "", history + [[user_message, None]]


    def bot(history, max_new_tokens, temperature, top_p, top_k, system_prompt):
        conv = get_conv_template("twllm_v2").copy()
        roles = {"human": conv.roles[0], "gpt": conv.roles[1]}  # map human to USER and gpt to ASSISTANT
        conv.system = system_prompt
        for user, bot in history:
            conv.append_message(roles['human'], user)
            conv.append_message(roles["gpt"], bot)
        msg = conv.get_prompt()
        prompt_tokens = tokenizer.encode(msg)
        length_of_prompt = len(prompt_tokens)
        if length_of_prompt > max_prompt_length:
            msg = tokenizer.decode(prompt_tokens[-max_prompt_length + 1:])

        history[-1][1] = ""
        for response in client.generate_stream(
                msg,
                max_new_tokens=max_new_tokens,
                temperature=temperature,
                top_p=top_p,
                top_k=top_k,
                repetition_penalty=1.1,
        ):
            if not response.token.special:
                character = response.token.text
                history[-1][1] += character
                yield history

        # After generating the response, store the conversation history in MongoDB
        conversation_document = {
            "model_name": model_name,
            "history": history,
            "system_prompt": system_prompt,
            "max_new_tokens": max_new_tokens,
            "temperature": temperature,
            "top_p": top_p,
            "top_k": top_k,
        }
        conversations_collection.insert_one(conversation_document)

    msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False).then(
        fn=bot,
        inputs=[
            chatbot,
            max_new_tokens,
            temperature,
            top_p,
            top_k,
            system_prompt,
        ],
        outputs=chatbot
    )
    submit_button.click(
        user, [msg, chatbot], [msg, chatbot], queue=False
    ).then(
        fn=bot,
        inputs=[
            chatbot,
            max_new_tokens,
            temperature,
            top_p,
            top_k,
            system_prompt,
        ],
        outputs=chatbot
    )


    def delete_prev_fn(
            history: list[tuple[str, str]]) -> tuple[list[tuple[str, str]], str]:
        try:
            message, _ = history.pop()
        except IndexError:
            message = ''
        return history, message or ''


    def display_input(message: str,
                      history: list[tuple[str, str]]) -> list[tuple[str, str]]:
        history.append((message, ''))
        return history

    retry_button.click(
        fn=delete_prev_fn,
        inputs=chatbot,
        outputs=[chatbot, saved_input],
        api_name=False,
        queue=False,
    ).then(
        fn=display_input,
        inputs=[saved_input, chatbot],
        outputs=chatbot,
        api_name=False,
        queue=False,
    ).then(
        fn=bot,
        inputs=[
            chatbot,
            max_new_tokens,
            temperature,
            top_p,
            top_k,
            system_prompt,
        ],
        outputs=chatbot,
    )

    undo_button.click(
        fn=delete_prev_fn,
        inputs=chatbot,
        outputs=[chatbot, saved_input],
        api_name=False,
        queue=False,
    ).then(
        fn=lambda x: x,
        inputs=[saved_input],
        outputs=msg,
        api_name=False,
        queue=False,
    )

    clear.click(lambda: None, None, chatbot, queue=False)

    gr.Markdown(LICENSE)

demo.queue(concurrency_count=4, max_size=128)
demo.launch()