lxml.html adds a find_class method to elements::
>>> from lxml.etree import Comment
>>> from lxml.html import document_fromstring, fragment_fromstring, tostring
>>> from lxml.html.clean import clean, clean_html
>>> from lxml.html import usedoctest
>>> try: unicode = __builtins__["unicode"]
... except (KeyError, NameError): unicode = str
>>> h = document_fromstring('''
...
...
... P1
... P2
... ''')
>>> print(tostring(h, encoding=unicode))
P1
P2
>>> print([e.text for e in h.find_class('fn')])
['P1']
>>> print([e.text for e in h.find_class('vcard')])
['P1', 'P2']
Also added is a get_rel_links, which you can use to search for links
like ````::
>>> h = document_fromstring('''
... test 1
... item 2
... item 3
... item 4''')
>>> print([e.attrib['href'] for e in h.find_rel_links('tag')])
['2', '4']
>>> print([e.attrib['href'] for e in h.find_rel_links('nofollow')])
[]
Another method is ``get_element_by_id`` that does what it says::
>>> print(tostring(fragment_fromstring('''
...
... stuff
...
''').get_element_by_id('test'), encoding=unicode))
stuff
Or to get the content of an element without the tags, use text_content()::
>>> el = fragment_fromstring('''
... ''')
>>> el.text_content()
'This is a bold link'
Or drop an element (leaving its content) or the entire tree, like::
>>> doc = document_fromstring('''
...
...
...
... This is a
test of stuff.
...
footer
...
... ''')
>>> doc.get_element_by_id('link').drop_tag()
>>> print(tostring(doc, encoding=unicode))
This is a test of stuff.
footer
>>> doc.get_element_by_id('body').drop_tree()
>>> print(tostring(doc, encoding=unicode))
footer
Note, however, that comment text will not be merged into the tree when you
drop the comment. Here, ``drop_tag()`` behaves exactly like ``drop_tree()``:
>>> for comment in doc.getiterator(Comment):
... comment.drop_tag()
>>> print(tostring(doc, encoding=unicode))
footer