import gradio as gr
import json
import spacy
import re
import string
import pandas as pd
import os
import requests
from textwrap import wrap

os.system('python -m spacy download en_core_web_sm')

nlp = spacy.load("en_core_web_sm")
nlp.add_pipe('sentencizer')


def download_and_save_file(URL, audio_dir):
    headers = {
        'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/72.0.3626.121 Safari/537.36',
        'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
        'referer': 'https://www.google.com/',
        'accept-encoding': 'gzip, deflate, br',
        'accept-language': 'en-US,en;q=0.9,',
        'cookie': 'prov=6bb44cc9-dfe4-1b95-a65d-5250b3b4c9fb; _ga=GA1.2.1363624981.1550767314; __qca=P0-1074700243-1550767314392; notice-ctt=4%3B1550784035760; _gid=GA1.2.1415061800.1552935051; acct=t=4CnQ70qSwPMzOe6jigQlAR28TSW%2fMxzx&s=32zlYt1%2b3TBwWVaCHxH%2bl5aDhLjmq4Xr',
    }
    doc = requests.get(URL, headers=headers)
    file_name = URL.split('/')[-1].split('?')[0]
    audio_path = f'{audio_dir}/{file_name}'
    with open(audio_path, 'wb') as f:
        f.write(doc.content)  
    return audio_path



credentials = os.environ['CREDENTIALS']
data = json.loads(credentials, strict=False)
with open('credentials.json', 'w') as f:
    json.dump(data, f)


gc = gspread.service_account(filename='credentials.json')
sh = gc.open('Annotated CC Audio')
worksheet = sh.sheet1
df = pd.DataFrame(worksheet.get_all_records())
sample_df = df[df['caption']==''].sample(1)

title = '🎵 Annotate audio'
description = '''Choose a sentence that describes audio the best if there's no such sentence please choose `No audio description`'''

audio_dir = 'AUDIO'
os.makedirs(audio_dir, exist_ok=True)

audio_id, audio_url, full_text, _ = sample_df.values[0]
audio_path = download_and_save_file(audio_url, audio_dir)
full_text = full_text.translate(str.maketrans('', '', string.punctuation))
sents = ['\n'.join(wrap(re.sub(r'###audio###\d###', '', s.text), width=70) )for s in nlp(full_text).sents]
sents.append('No audio description')

def audio_demo(cap, audio, audio_id):

    df.at[int(audio_id)-1, 'caption'] = cap
    worksheet.update([df.columns.values.tolist()] + df.values.tolist())
    return 'success!'


iface = gr.Interface(
    audio_demo,  
    inputs=[gr.Dropdown(sents, label='audio description'), gr.Audio(audio_path, type="filepath"), gr.Textbox(value=audio_id, visible=False)],
    outputs=[gr.Textbox(label="output")],
    allow_flagging="never",
    title=title,
    description=description,
    )

if __name__ == "__main__":
  iface.launch(show_error=True, debug=True)