HTML to JSON

gistfile1.py

#!/usr/bin/env python
from bs4 import BeautifulSoup
import json

data = {'photos': []}
soup = BeautifulSoup(open('gallery.html'))

for item in soup.find_all('li'):
    if 'testimony' in item['class']:
        continue
    obj = {
        'description': item.a['data-description'],
        'featured': 'box-6' in item['class'],
        'href': item.a['href'],
        'img': item.a.img['src'],
        'tags': [tag for tag in item['class'] if tag not in ('box-3', 'box-6', 'is-loading')],
        'title': item.a['title']
    }
    data['photos'].append(obj)

with open('gallery.json', 'w') as fp:
    json.dump(data, fp, sort_keys=True, indent=4)