Cleaning up HTML snippets

from elementtidy.TidyHTMLTreeBuilder import TidyHTMLTreeBuilder
from elementtree.ElementTree import tostring

# list of accepted tags
VALID_TAGS = (
    "a", "p", "blockquote", "ul", "ol", "li", "dl", "dt", "dd", "em",
    "strong", "dfn", "code", "q", "samp", "kbd", "var", "cite",
    "abbr", "acronym", "sub", "sup"
    )

# dictionary mapping presentational tags to semantic tags
TAG_MAP = { "b": "strong", "i": "em", "tt": "samp" }

# XHTML namespace url
NS_XHTML = "{http://www.w3.org/1999/xhtml}"

def fixsnippet(snippet):

    parser = TidyHTMLTreeBuilder()
    parser.feed(snippet)

    body = parser.close().find(NS_XHTML + "body")

    for elem in body:
        for elem in elem.getiterator():
            # get rid of XHTML namespace
            if elem.tag.startswith(NS_XHTML):
                elem.tag = elem.tag[len(NS_XHTML):]

            # map presentation tags to semantic tags
            elem.tag = TAG_MAP.get(elem.tag, elem.tag)

            # check for bad tags
            if elem.tag not in VALID_TAGS:
                raise SyntaxError("invalid tag: %r" % elem.tag)

            # clear out attributes (extend as necessary)
            if elem.tag == "a":
                href = elem.get("href")
                elem.attrib.clear()
                if href:
                    elem.set("href", href)
            else:
                elem.attrib.clear()

    # change the body tag to a div tag
    body.tag = "div"; body.attrib.clear()
    return tostring(body)