etree treewalker infinite loop

[gsnedders]

This goes into an infinite loop:

import html5lib

frag = html5lib.parseFragment("<b><em><foo><foob><fooc><aside></b></em>")

walker = html5lib.getTreeWalker("etree")

print list(walker(frag))

[gsnedders]

This is weird.

(Pdb) print list(frag)
[<Element u'{http://www.w3.org/1999/xhtml}b' at 0x7fdd8eb42990>, <Element u'{http://www.w3.org/1999/xhtml}aside' at 0x7fdd8eb42e10>, <Element u'{http://www.w3.org/1999/xhtml}aside' at 0x7fdd8eb42e10>]

Note it's a fragment containing the same element twice.

[gsnedders]

It's still strictly a tree, though, so I don't think the tree walker should break so badly?

[gsnedders]

list(parents[-1]).index(parent) in the etree walker is why. The above commit adds an assertion that the count is 1, which at least prevents the infinite loop.