import gradio as gr
import requests
from pypdf import PdfReader

def scrape(instring):
    html_src=(f'''
    <div style="text-align:center">
    <h4>PDF Viewer</h4>
    <iframe src="https://docs.google.com/viewer?url={instring}&embedded=true" frameborder="0" height="1200px" width="100%"></iframe>
    </div>''')
    return gr.HTML.update(f'''{html_src}''')

def scrape00(instring, page_num):
    response = requests.get(instring, stream=True)

    if response.status_code == 200:
        with open("data.pdf", "wb") as f:
            f.write(response.content)
    else:
        print(response.status_code)


    #out = Path("./data.pdf")
    #print (out)
    reader = PdfReader("data.pdf")
    number_of_pages = len(reader.pages)
    page = reader.pages[int(page_num)-1]
    text = page.extract_text()
    print (text)
    try:
        summarizer = gr.Interface.load("huggingface/facebook/bart-large-cnn")
        sum_out = summarizer(text)
    except Exception:
        sum_out = "Error"
    return text, sum_out

with gr.Blocks() as app:
    gr.Markdown('''<h1>PDF Viewer''')
    with gr.Row():
        inp=gr.Textbox(label="PDF URL",scale=3)
        pg_num=gr.Number(label="Page Number",value=1,precision=0,scale=1)
    with gr.Row():
        go_btn = gr.Button("Load PDF")
        sum_btn = gr.Button("Summarize")
    outp = gr.HTML()
    with gr.Row():
        text_out = gr.Textbox()
        sum_out = gr.Textbox()
    go_btn.click(scrape,inp,outp)
    sum_btn.click(scrape00,[inp,pg_num],[text_out,sum_out])
app.queue(concurrency_count=10).launch()