1 # -*- coding: utf-8 -*-
2 # CurlyTx Atom feed parser
3 # Copyright (C) 2011 Christian Weiske <cweiske@cweiske.de>
4 # License: GPLv3 or later
6 from twisted.web.client import getPage
7 from xml.etree.cElementTree import fromstring
10 """ Simple XML parser that extracts pages from a atom feed """
11 ns = "{http://www.w3.org/2005/Atom}"
12 def __init__(self, url, callback):
15 Parsed pages are sent back to callback by parse()
17 getPage(url).addCallback(self.parse, callback).addErrback(self.fail)
23 def parse(self, data, callback):
24 """ Parse atom feed data into pages list and run callback """
25 xml = fromstring(data)
27 for entry in xml.findall("{0}entry".format(self.ns)):
28 titleE = entry.find("{0}title".format(self.ns))
29 url = self.bestLink(entry.findall("{0}link".format(self.ns)))
30 if titleE != None and titleE.text != "" and url != None:
31 pages.append({"title": titleE.text, "url": url})
35 def bestLink(self, list):
36 """ Fetch the best matching link from an atom feed entry """
40 if link.get("rel") != "alternate" and link.get("rel") != "":
42 level = self.level(link)
43 if foundLevel > level:
46 foundHref = link.get("href")
49 def level(self, link):
50 """ Determines the level of a link
52 "text/plain" type links are best, links without type are second.
53 All others have the lowest level 1.
55 type = link.get("type")
56 if type == "text/plain":