html5lib_lxml_walker_bug.py

#!/usr/bin/env python3

import html5lib

def test_parse_serialize(html, tree_type):
  walker = html5lib.getTreeWalker(tree_type)
  serializer = html5lib.serializer.HTMLSerializer()
  #root = html5lib.parseFragment(html, namespaceHTMLElements=False, treebuilder=tree_type)
  root = html5lib.parseFragment(html, treebuilder=tree_type)
  stream = walker(root)
  return serializer.render(stream)

def test_html(html, tree_types=('etree', 'lxml')):
  print ('\ntesting %r:' % html)
  for tt in ('etree', 'lxml'):
    print (' with %r:\n  %r' % (tt, test_parse_serialize(html, tt)))

if __name__ == '__main__':
  # works
  test_html('<a>one</a>')
  # errors out for 'lxml' ('etree' works fine)
  test_html('')