from lxml import html, etree
html_str = """
this is my div
"""
#parsing html string with lxml.html is less restrictive than lxml.etree
#doc = etree.fromstring(html_str)
doc = html.fromstring(html_str)
items = doc.xpath('//div[@id="my_div"]')
my_div = items[0]
print my_div.tag
print my_div.text
print my_div.attrib
items = doc.xpath('/html/body')
body = items[0]
ul = etree.SubElement(body, 'ul')
li = etree.SubElement(ul, 'li')
a = etree.SubElement(li, 'a', href = 'http://infohost.nmt.edu/tcc/help/pubs/pylxml/pylxml.pdf')
a.text = "Lxml doc I"
span = etree.SubElement(a, 'span')
span.set('class','caret')
ul = etree.SubElement(body, 'ul')
li = etree.SubElement(ul, 'li')
a = etree.SubElement(li, 'a', href = 'http://lxml.de/')
a.text = "Lxml doc II"
span = etree.SubElement(a, 'span')
span.set('class','caret')
print html.tostring(doc)