#!/usr/bin/python

from rfc822 import parseaddr
from post import sanitize
import atomef

from lazydom import namespace
atom=namespace('http://purl.org/atom/ns#')
dc=namespace('http://purl.org/dc/elements/1.1/')
content=namespace('http://purl.org/rss/1.0/modules/content/')
http=namespace('http://schemas.xmlsoap.org/wsdl/http/')
rdf=namespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#')
rss=namespace('http://purl.org/rss/1.0/')
soap=namespace('http://schemas.xmlsoap.org/soap/envelope/')
wfw=namespace('http://wellformedweb.org/CommentAPI/')
xhtml=namespace('http://www.w3.org/1999/xhtml')

oktags = ['a','b', 'br', 'img', 'p', 'strong', 'em', 'ul', 'li', 'blockquote']

def validate(xml):
  # http://diveintomark.org/archives/2003/06/12/how_to_consume_rss_safely
  for element in xml[:]:
    if not element.name[1] in oktags:
      class DisallowedTag(Exception): pass
      raise DisallowedTag, "Disallowed tag: " + element.name[1]
    if 'style' in map(unicode.upper,element.node.attributes.keys()):
      class DisallowedAttr(Exception): pass
      raise DisallowedAttr, "Disallowed attribute: style"
    validate(element)

def parse(entry):
  entry = atomef.parse(entry)

  # soap Headers
  for header in entry[soap.Header][:]:
    if header[soap._mustUnderstand]=='1': 
      if header.name == wfw.preview:
        doPreview=True
      else:
        class NotUnderstood(Exception): pass
        raise NotUnderstood, "Header %s not understood" % str(header.node.nodeName)
    del header
  if entry[soap.Body]: entry=entry[soap.Body][:]; addSoap=True
  if entry.name == http.POST: entry=entry[:]

  # rdf Marker
  if entry.name==rdf.RDF: entry=entry[:]; addRdf=True

  # title
  title = entry[atom.title] or entry['title'] or entry[rss.title]

  # content
  content=globals()['content']
  body = entry[atom.content] or entry[atom.summary]
  body = body or entry[xhtml.body] or entry[content.encoded]
  body = body or entry[rss.description] or entry['description']

  if body and body.name[1] in ['description','encoded']:
    body.node.setAttribute('mode','escaped')
    body.node.setAttribute('type','text/html')
  if body and body.name == xhtml.body:
    body.node.setAttribute('type','text/html')

  if len(body[:])==1 and body[:].name[1] in ['div','span']:
    if body[:].node.attributes.keys() in [[],['xmlns']]: 
      child=body[:].node
      for ((ns,qname),name) in body.node.attributes.itemsNS():
        child.setAttributeNS(ns,qname,name)
      del child,ns,qname,name
      body=body[:]

  validate(body)
  content = body.toHtml()

  if body["@mode"] == 'escaped': 
    content = sanitize(content)
  del body

  # determine email
  email = entry[atom.author][atom.email] or entry['email']

  # determine link
  link = entry[atom.link] or entry[atom.author][atom.url]
  link = link or entry[rss.link] or entry['link'] or entry[rdf._about]
  if not link and email: link = "mailto:" + str(email)

  # determine name
  name = entry[atom.author][atom.name] or entry[dc.creator] or entry['author']
  if name.toString().find('@')>0: (name, email)=parseaddr(name.toString())
  if not name and link and str(link).startswith('http://'):
    try:
      from pingback import parser
      file=urllib.urlopen(comment.link.split('#')[0])
      page = parser()
      page.feed(file.read())
      file.close()
      name = page.title
      del parser, file, page
    except: pass
  if not name: name="anonymous"

  # issued
  from template import parseDate
  issued = parseDate(str(entry[atom.issued] or entry[dc.date]))

  del entry
  return locals()