"""Stable identities from publisher-owned metadata self-links.""" import re from urllib.parse import urlsplit def mlsys_id(url): parsed = urlsplit(url) if parsed.scheme not in ('https', 'http') or parsed.netloc != 'proceedings.mlsys.org': return None match = re.fullmatch(r'/paper(?:_files/paper)?/((?:19|20)\d{2})/hash/([a-f0-9]{32})-Abstract(?:-(?:Conference|mlsys(20\d{2})))?\.html', parsed.path) if not match or match[3] and match[3] != match[1]: return None return match[1] + '/' + match[2]