+from HTMLParser import HTMLParser
+
+class HTML2Text(HTMLParser):
+ entities = {
+ "amp": "&",
+ "lt": "<",
+ "gt": ">",
+ "pound": "£",
+ "copy": "©",
+ "apos": "'",
+ "quot": "\"",
+ "nbsp": " ",
+ }
+
+ blockleveltags = [
+ "h1",
+ "h2",
+ "h3",
+ "h4",
+ "h5",
+ "h6",
+ "pre",
+ "p",
+ "ul",
+ "ol",
+ "dl",
+ "br",
+ ]
+
+ liststarttags = [
+ "ul",
+ "ol",
+ "dl",
+ ]
+
+ cancontainflow = [
+ "div",
+ "li",
+ "dd",
+ "blockquote",
+ ]
+
+ def __init__(self,textwidth=70):
+ self.text = u''
+ self.curdata = u''
+ self.textwidth = textwidth
+ self.opentags = []
+ self.indentlevel = 0
+ self.listcount = []
+ HTMLParser.__init__(self)
+
+ def handle_starttag(self, tag, attrs):
+ tag_name = tag.lower()
+ if tag_name in self.blockleveltags:
+ # handle starting a new block - unless we're in a block element
+ # that can contain other blocks, we'll assume that we want to close
+ # the container
+ if tag_name == u'br':
+ self.handle_curdata()
+ self.opentags.append(tag_name)
+ self.opentags.pop()
+
+ if tag_name == u'ol':
+ self.handle_curdata()
+ self.listcount.append(1)
+ self.listlevel = len(self.listcount) - 1
+
+ if tag_name in self.liststarttags:
+ smallist = self.opentags[-3:]
+ smallist.reverse()
+ for prev_listtag in smallist:
+ if prev_listtag in [u'dl', u'ol']:
+ self.indentlevel = self.indentlevel + 4
+ break
+ elif prev_listtag == u'ul':
+ self.indentlevel = self.indentlevel + 3
+ break
+
+ if len(self.opentags) > 0:
+ self.handle_curdata()
+ if tag_name not in self.cancontainflow:
+ self.opentags.pop()
+ self.opentags.append(tag_name)
+ else:
+ listcount = 0
+ try:
+ listcount = self.listcount[-1]
+ except:
+ pass
+ self.handle_curdata()
+ self.opentags.append(tag_name)
+
+ def handle_startendtag(self, tag, attrs):
+ if tag.lower() == u'br':
+ self.tags.append(u'br')
+ self.handle_curdata() # just handle the data, don't do anything else
+ self.tags.pop()
+
+ def handle_curdata(self):
+ if len(self.opentags) == 0:
+ return
+
+ if len(self.curdata) == 0:
+ return
+
+ if len(self.curdata.strip()) == 0:
+ return
+
+ tag_thats_done = self.opentags[-1]
+
+ if tag_thats_done in self.blockleveltags:
+ newlinerequired = self.text != u''
+ if newlinerequired:
+ if newlinerequired \
+ and len(self.text) > 2 \
+ and self.text[-1] != u'\n' \
+ and self.text[-2] != u'\n':
+ self.text = self.text + u'\n\n'
+
+ if tag_thats_done in ["h1", "h2", "h3", "h4", "h5", "h6"]:
+ underline = u''
+ underlinechar = u'='
+ headingtext = self.curdata.encode("utf-8").strip()
+ seperator = u'\n' + u' '*self.indentlevel
+ headingtext = seperator.join( \
+ textwrap.wrap( \
+ headingtext, \
+ self.textwidth - self.indentlevel \
+ ) \
+ )
+
+ if tag_thats_done == u'h2':
+ underlinechar = u'-'
+ elif tag_thats_done != u'h1':
+ underlinechar = u'~'
+
+ if u'\n' in headingtext:
+ underline = u' ' * self.indentlevel \
+ + underlinechar * (self.textwidth - self.indentlevel)
+ else:
+ underline = u' ' * self.indentlevel \
+ + underlinechar * len(headingtext)
+ self.text = self.text \
+ + headingtext.encode("utf-8") + u'\n' \
+ + underline
+ elif tag_thats_done == "p":
+ paragraph = self.curdata.encode("utf-8").strip()
+ seperator = u'\n' + u' ' * self.indentlevel
+ self.text = self.text \
+ + u' ' * self.indentlevel \
+ + seperator.join(textwrap.wrap(paragraph, self.textwidth - self.indentlevel))
+ elif tag_thats_done == "pre":
+ self.text = self.text + self.curdata
+ elif tag_thats_done == "blockquote":
+ quote = self.curdata.encode("utf-8").strip()
+ seperator = u'\n' + u' ' * self.indentlevel + u'> '
+ self.text = self.text \
+ + u'> ' \
+ + seperator.join( \
+ textwrap.wrap( \
+ quote, \
+ self.textwidth - self.indentlevel - 2 \
+ )
+ )
+ elif tag_thats_done == "li":
+ item = self.curdata.encode("utf-8").strip()
+ if len(self.text) > 0 and self.text[-1] != u'\n':
+ self.text = self.text + u'\n'
+ # work out if we're in an ol rather than a ul
+ latesttags = self.opentags[-4:]
+ latesttags.reverse()
+ isul = False
+ for thing in latesttags:
+ if thing == 'ul':
+ isul = True
+ break
+ elif thing == 'ol':
+ isul = False
+ break
+
+ listindent = 3
+ if not isul:
+ listindent = 4
+
+ listmarker = u' * '
+ if not isul:
+ listmarker = u' %2d. ' %(self.listcount[-1])
+ self.listcount[-1] = self.listcount[-1] + 1
+
+ seperator = u'\n' \
+ + u' ' * self.indentlevel \
+ + u' ' * listindent
+ self.text = self.text \
+ + u' ' * self.indentlevel \
+ + listmarker \
+ + seperator.join( \
+ textwrap.wrap( \
+ item, \
+ self.textwidth - self.indentlevel - listindent \
+ ) \
+ )
+ self.curdata = u''
+ elif tag_thats_done == "dt":
+ definition = self.curdata.encode("utf-8").strip()
+ if len(self.text) > 0 and self.text[-1] != u'\n':
+ self.text = self.text + u'\n\n'
+ elif len(self.text) > 1 and self.text[-2] != u'\n':
+ self.text = self.text + u'\n'
+ definition = definition + "::"
+ self.text = self.text \
+ + '\n '.join(
+ textwrap.wrap(definition, self.textwidth - 1))
+ self.curdata = u''
+ elif tag_thats_done == "dd":
+ definition = self.curdata.encode("utf-8").strip()
+ if len(definition) > 0:
+ if len(self.text) > 0 and self.text[-1] != u'\n':
+ self.text = self.text + u'\n'
+ self.text = self.text \
+ + ' ' \
+ + '\n '.join( \
+ textwrap.wrap( \
+ definition, \
+ self.textwidth - self.indentlevel - 4 \
+ ) \
+ )
+ self.curdata = u''
+ elif tag_thats_done in self.liststarttags:
+ pass
+ else:
+ # we've got no idea what this tag does, so we'll
+ # make an assumption that we're not going to know later
+ if len(self.curdata) > 0:
+ self.text = self.text \
+ + u' ... ' \
+ + u'\n ... '.join( \
+ textwrap.wrap(self.curdata, self.textwidth - 5))
+ self.curdata = u''
+
+ if tag_thats_done in self.blockleveltags:
+ self.curdata = u''
+
+ def handle_endtag(self, tag):
+ try:
+ tagindex = self.opentags.index(tag)
+ except:
+ # closing tag we know nothing about.
+ # err. weird.
+ tagindex = 0
+
+ tag = tag.lower()
+
+ if tag in self.liststarttags:
+ if tag in [u'ol', u'dl', u'ul']:
+ # find if there was a previous list level
+ smalllist = self.opentags[:-1]
+ smalllist.reverse()
+ for prev_listtag in smalllist:
+ if prev_listtag in [u'ol', u'dl']:
+ self.indentlevel = self.indentlevel - 4
+ break
+ elif prev_listtag == u'ul':
+ self.indentlevel = self.indentlevel - 3
+ break
+
+ if tag == u'ol':
+ self.listcount = self.listcount[:-1]
+
+ while tagindex < len(self.opentags) \
+ and tag in self.opentags[tagindex+1:]:
+ try:
+ tagindex = self.opentags.index(tag, tagindex+1)
+ except:
+ # well, we don't want to do that then
+ pass
+ if tagindex != len(self.opentags) - 1:
+ # Assuming the data was for the last opened tag first
+ self.handle_curdata()
+ # Now kill the list to be a slice before this tag was opened
+ self.opentags = self.opentags[:tagindex + 1]
+ else:
+ self.handle_curdata()
+ if self.opentags[-1] == tag:
+ self.opentags.pop()
+
+ def handle_data(self, data):
+ self.curdata = self.curdata + unicode(data, "utf-8")
+
+ def handle_entityref(self, name):
+ entity = name
+ if HTML2Text.entities.has_key(name.lower()):
+ entity = HTML2Text.entities[name.lower()]
+ elif name[0] == "#":
+ entity = unichr(int(name[1:]))
+ else:
+ entity = "&" + name + ";"
+
+ self.curdata = self.curdata + unicode(entity, "utf-8")
+
+ def gettext(self):
+ self.handle_curdata()
+ if len(self.text) == 0 or self.text[-1] != u'\n':
+ self.text = self.text + u'\n'
+ self.opentags = []
+ if len(self.text) > 0:
+ while len(self.text) > 1 and self.text[-1] == u'\n':
+ self.text = self.text[:-1]
+ self.text = self.text + u'\n'
+ return self.text
+
+def open_url(method, url):
+ redirectcount = 0
+ while redirectcount < 3:
+ (type, rest) = urllib.splittype(url)
+ (host, path) = urllib.splithost(rest)
+ (host, port) = urllib.splitport(host)
+ if port == None:
+ port = 80
+ try:
+ conn = httplib.HTTPConnection("%s:%s" %(host, port))
+ conn.request(method, path)
+ response = conn.getresponse()
+ if response.status in [301, 302, 303, 307]:
+ headers = response.getheaders()
+ for header in headers:
+ if header[0] == "location":
+ url = header[1]
+ elif response.status == 200:
+ return response
+ except:
+ pass
+ redirectcount = redirectcount + 1
+ return None
projects / rss2maildir.git / blobdiff