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, ParseError
12 """ Simple XML parser that extracts pages from a atom feed """
13 ns = "{http://www.w3.org/2005/Atom}"
16 def __init__(self, url, callback, errorCallback):
19 Parsed pages are sent back to callback by parse()
21 self.errorCallback = errorCallback
22 if (url.startswith('file://')):
24 if not os.path.exists(file):
25 errorCallback('Settings atom feed file does not exist: ' + file)
28 with open(file, 'r') as f:
29 self.parse(f.read(), callback)
31 getPage(url).addCallback(self.parse, callback).addErrback(self.onError)
33 def parse(self, data, callback):
34 """ Parse atom feed data into pages list and run callback """
36 xml = fromstring(data)
38 return self.errorCallback("Invalid XML")
41 for entry in xml.findall("{0}entry".format(self.ns)):
42 titleE = entry.find("{0}title".format(self.ns))
43 url = self.bestLink(entry.findall("{0}link".format(self.ns)))
44 if titleE != None and titleE.text != "" and url != None:
45 pages.append({"title": titleE.text, "url": url})
49 def bestLink(self, list):
50 """ Fetch the best matching link from an atom feed entry """
54 if link.get("rel") != "alternate" and link.get("rel") != "":
56 level = self.level(link)
57 if foundLevel > level:
60 foundHref = link.get("href")
63 def level(self, link):
64 """ Determines the level of a link
66 "text/plain" type links are best, links without type are second.
67 All others have the lowest level 1.
69 type = link.get("type")
70 if type == "text/plain":
76 def onError(self, error):
77 """ Pass the error message only """
78 self.errorCallback(error.getErrorMessage())