import gradio as gr
from openai import OpenAI
import os
from fpdf import FPDF  # For PDF conversion
from docx import Document  # For DOCX conversion

css = '''
.gradio-container{max-width: 1000px !important}
h1{text-align:center}
footer {
    visibility: hidden
}
'''

ACCESS_TOKEN = os.getenv("HF_TOKEN")

client = OpenAI(
    base_url="https://api-inference.huggingface.co/v1/",
    api_key=ACCESS_TOKEN,
)

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    messages = [{"role": "system", "content": system_message}]

    for val in history:
        if val[0]:
            messages.append({"role": "user", "content": val[0]})
        if val[1]:
            messages.append({"role": "assistant", "content": val[1]})

    messages.append({"role": "user", "content": message})

    response = ""
    
    for message in  client.chat.completions.create(
        model="meta-llama/Meta-Llama-3.1-8B-Instruct",
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
        messages=messages,
    ):
        token = message.choices[0].delta.content
        
        response += token
        yield response

def save_as_file(input_text, output_text, conversion_type):
    if conversion_type == "PDF":
        pdf = FPDF()
        pdf.add_page()
        pdf.set_font("Arial", size=12)
        pdf.multi_cell(0, 10, f"User Query: {input_text}\n\nResponse: {output_text}")
        file_name = "output.pdf"
        pdf.output(file_name)
    elif conversion_type == "DOCX":
        doc = Document()
        doc.add_heading('Conversation', 0)
        doc.add_paragraph(f"User Query: {input_text}\n\nResponse: {output_text}")
        file_name = "output.docx"
        doc.save(file_name)
    elif conversion_type == "TXT":
        file_name = "output.txt"
        with open(file_name, "w") as f:
            f.write(f"User Query: {input_text}\n\nResponse: {output_text}")
    else:
        return None

    return file_name

def convert_and_download(history, conversion_type):
    if not history:
        return None
    
    input_text = "\n".join([f"User: {h[0]}" for h in history if h[0]])
    output_text = "\n".join([f"Assistant: {h[1]}" for h in history if h[1]])

    file_path = save_as_file(input_text, output_text, conversion_type)
    return file_path

def handle_conversion(history, conversion_type):
    file_path = convert_and_download(history, conversion_type)
    return gr.File(file_path)

demo = gr.Blocks(css=css)

with demo:
    with gr.Row():
        system_message = gr.Textbox(value="", label="System message")
        max_tokens = gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens")
        temperature = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature")
        top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-P")
    
    history = gr.State(value=[])
    
    chatbot = gr.ChatInterface(
        fn=respond,
        additional_inputs=[system_message, max_tokens, temperature, top_p],
    )

    with gr.Row():
        conversion_type = gr.Dropdown(choices=["PDF", "DOCX", "TXT"], label="Conversion Type")
        download_button = gr.Button("Convert and Download")
    
    file_output = gr.File()
    
    download_button.click(
        handle_conversion,
        inputs=[history, conversion_type],
        outputs=file_output
    )

if __name__ == "__main__":
    demo.launch()