import patch22
import lazydom
from htmlentitydefs import name2codepoint, codepoint2name

import re
from xml.dom import minidom,Node
from sgmllib import SGMLParser

def unescape(data):
  "convert html entitydefs into unicode characters"
  chunks = re.split('&(#?\w+);',data)
  for i in range(1,len(chunks),2):
    if chunks[i] in name2codepoint:
      chunks[i] = unichr(name2codepoint[chunks[i]])
    elif re.match('#\d+$',chunks[i]):
      chunks[i] = unichr(int(chunks[i][1:]))
  return "".join(chunks)

class stripHtml(SGMLParser):
  "remove all tags from the data"
  def __init__(self, data):
    SGMLParser.__init__(self)
    self.result=''
    self.feed(data)
    self.close()
  def handle_data(self, data):
    if data: self.result+=data

def escape(data,entity=1):
  "escape all entities and non-7bit ASCII characters"
  for i in range(len(data)-1,-1,-1):
    n=ord(data[i])
    if (n<128) and not entity: continue
    if n in codepoint2name:
      data = '%s&%s;%s' % (data[:i], codepoint2name[n], data[i+1:])
    elif n>=128:
      data = '%s&#%d;%s' % (data[:i], n, data[i+1:])
  return data

def parse(stream):
  "parse a stream into atomef"
  return atomef(minidom.parse(stream).documentElement)

class atomef(lazydom.lazydom):
  def __init__(self,element=None):
    lazydom.lazydom.__init__(self,element)

  def toHtml(self):

    if self['@mode'] == 'escaped':
      data=''.join([n.nodeValue for n in self.list[0].childNodes
        if n.nodeType==Node.TEXT_NODE or n.nodeType == Node.CDATA_SECTION_NODE])
    else:
      data=''.join([n.toxml() for n in self.list[0].childNodes])

    if self['@type'] <> 'text/html':
      data = escape(data,1)
    else:
      data = escape(data,0)

    return data

  def toString(self):

    data=''.join([n.nodeValue for n in self.list[0].childNodes
      if n.nodeType==Node.TEXT_NODE or n.nodeType == Node.CDATA_SECTION_NODE])

    if self['@mode'] == 'escaped':
      data = unescape(data)

    if self['@type'] == 'text/html':
      data = stripHtml(data).result

    return data

if __name__ == '__main__':
  print unescape('&lt;&#62;')
  print escape(unescape('Se&ntilde;or&nbsp;Gonz&aacute;lez'))
  print stripHtml(unescape('not &lt;b&gt;very&lt;/b&gt; nice')).result
  print escape(unescape('&#128;'))