Seems to be hit quite often, so make it helpful.
# comes from parse_id()[0]. See the
# contents_anchor_in_both_group_and_namespace test for a verification.
id = element.attrib['id']
- assert id.startswith(state.current_definition_url_base)
+ assert id.startswith(state.current_definition_url_base), "ID `%s` does not start with `%s`" % (id, state.current_definition_url_base)
return id[len(state.current_definition_url_base)+2:]
and_re_src = re.compile(r'([^\s])&&([^\s])')