extruct-test.py

gistfile1.txt

import pprint

import extruct
import requests
from w3lib.html import get_base_url

# https://www.crosswaterlondon.com/product/mpro-towel-holder?variant=54381#image-2
pp = pprint.PrettyPrinter(indent=0)
r = requests.get('https://www.signaturehardware.com/ruscello-widespread-bathroom-faucet/951334.html')
base_url = get_base_url(r.text, r.url)
data = extruct.extract(r.text, base_url=base_url)
pp.pprint(data)