import gradio as gr
from functions import extract_text_from_pdf, format_content, split_into_snippets, build_prompts

def process_inputs(pdf_file, model_choice, output_format, oauth_token: gr.OAuthToken | None = None):
    """Process PDF and generate summary"""
    if oauth_token is None:
        return "### Please log in to use this service"
    
    if not pdf_file:
        return "### Please upload a PDF file"
    
    try:
        text = extract_text_from_pdf(pdf_file.name)
        return f"### Processing successful with {model_choice}!"
    except Exception as e:
        return f"### Error: {str(e)}"

# Define core interface components
iface = gr.Interface(
    fn=process_inputs,
    inputs=[
        gr.File(
            label="Upload PDF",
            file_types=[".pdf"]
        ),
        gr.Dropdown(
            choices=[
                "GPT-3.5",
                "GPT-4",
                "Claude-3",
                "Mistral"
            ],
            label="Model",
            value="GPT-3.5"
        ),
        gr.Radio(
            choices=["TXT", "MD", "HTML"],
            label="Format",
            value="TXT"
        )
    ],
    outputs=gr.Markdown(
        label="Output",
        value="### Upload your PDF to begin"
    ),
    flagging_mode="never",
    css="""
        .gradio-container {
            max-width: 800px !important;
            margin: 0 auto !important;
        }
        .container {
            max-width: 800px !important;
            margin: 0 auto !important;
            padding: 2rem !important;
        }
    """
)

# Create main app
with gr.Blocks(theme=gr.themes.Default()) as demo:
    gr.Markdown("## 🚀 PDF to LLM Summarizer")
    
    with gr.Row():
        with gr.Column():
            gr.Markdown("📄 Extract and summarize text from PDFs using state-of-the-art language models")
        with gr.Column():
            gr.LoginButton(min_width=200)
    
    iface.render()
    
    gr.Markdown("Made with Gradio")

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