import nltk
nltk.download('punkt')

def create_html(title, content):
  html = f"""<div style='max-width:100%; max-height:360px; overflow:auto'>
            <h1>{title}</h1>
            <br/>
            <p>{content}</p>
        </div>"""
  return html
    
import numpy as np
def get_max(data):
  return max(data, key=lambda x: x['score'])

import wikipediaapi
from sumy.parsers.plaintext import PlaintextParser
from sumy.nlp.tokenizers import Tokenizer
from sumy.summarizers.lsa import LsaSummarizer
def summarize_wikipedia(search_query, language='en', sentences_count=3):
    wiki_wiki = wikipediaapi.Wikipedia('AgungBagus (agungbagus@example.com)',language)
    page = wiki_wiki.page(search_query)

    if not page.exists():
        return "Article not found."

    content = page.text
    parser = PlaintextParser.from_string(content, Tokenizer(language))
    summarizer = LsaSummarizer()
    summary = summarizer(parser.document, sentences_count)

    return ' '.join([str(sentence) for sentence in summary])

import gradio as gr
from PIL import Image
from transformers import pipeline

def image_processing(input_image : gr.Image):
  image = Image.fromarray(input_image)
  classifier = pipeline(task="image-classification", model="gungbgs/bird_species_classifier")
  species = get_max(classifier(image))['label']
  species_text = species.lower()
  result = summarize_wikipedia(species_text)
  return create_html(species, result)


app = gr.Interface(
    fn = image_processing,
    inputs = "image",
    outputs = "html"
)

app.launch(share=True)