src/gpodder/opml.py

   1 # -*- coding: utf-8 -*-
   2 #
   3 # gPodder - A media aggregator and podcast client
   4 # Copyright (c) 2005-2008 Thomas Perl and the gPodder Team
   5 #
   6 # gPodder is free software; you can redistribute it and/or modify
   7 # it under the terms of the GNU General Public License as published by
   8 # the Free Software Foundation; either version 3 of the License, or
   9 # (at your option) any later version.
  10 #
  11 # gPodder is distributed in the hope that it will be useful,
  12 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 # GNU General Public License for more details.
  15 #
  16 # You should have received a copy of the GNU General Public License
  17 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
  18 #
  19
  20
  21 #
  22 #  opml.py -- OPML import and export functionality
  23 #  Thomas Perl <thp@perli.net>   2007-08-19
  24 #
  25 #  based on: libopmlreader.py (2006-06-13)
  26 #            libopmlwriter.py (2005-12-08)
  27 #
  28
  29 """OPML import and export functionality
  30
  31 This module contains helper classes to import subscriptions
  32 from OPML files on the web and to export a list of channel
  33 objects to valid OPML 1.1 files that can be used to backup
  34 or distribute gPodder's channel subscriptions.
  35 """
  36
  37 from gpodder.liblogger import log
  38
  39 from gpodder import util
  40
  41 import gtk
  42 import gobject
  43
  44 import xml.dom.minidom
  45 import xml.sax.saxutils
  46
  47 import urllib
  48 import urllib2
  49 import os.path
  50 import os
  51
  52 import datetime
  53 import gpodder
  54
  55
  56 class Importer(object):
  57     """
  58     Helper class to import an OPML feed from protocols
  59     supported by urllib2 (e.g. HTTP) and return a GTK
  60     ListStore that can be displayed in the GUI.
  61
  62     This class should support standard OPML feeds and
  63     contains workarounds to support odeo.com feeds.
  64     """
  65
  66     VALID_TYPES = ( 'rss', 'link' )
  67
  68     def read_url( self, url):
  69         request = urllib2.Request( url, headers = {'User-agent': gpodder.user_agent})
  70         return urllib2.urlopen( request).read()
  71
  72     def __init__( self, url):
  73         """
  74         Parses the OPML feed from the given URL into
  75         a local data structure containing channel metadata.
  76         """
  77         self.items = []
  78         try:
  79             if url.startswith('/'):
  80                 # assume local filename
  81                 if os.path.exists(url):
  82                     doc = xml.dom.minidom.parse( url)
  83                 else:
  84                     log('Empty/non-existing OPML file', sender=self)
  85                     return
  86             else:
  87                 doc = xml.dom.minidom.parseString( self.read_url( url))
  88
  89             for outline in doc.getElementsByTagName('outline'):
  90                 if outline.getAttribute('type') in self.VALID_TYPES and outline.getAttribute('xmlUrl') or outline.getAttribute('url'):
  91                     channel = {
  92                         'url': outline.getAttribute('xmlUrl') or outline.getAttribute('url'),
  93                         'title': outline.getAttribute('title') or outline.getAttribute('text') or outline.getAttribute('xmlUrl') or outline.getAttribute('url'),
  94                         'description': outline.getAttribute('text') or outline.getAttribute('xmlUrl') or outline.getAttribute('url'),
  95                     }
  96
  97                     if channel['description'] == channel['title']:
  98                         channel['description'] = channel['url']
  99
 100                     for attr in ( 'url', 'title', 'description' ):
 101                         channel[attr] = channel[attr].strip()
 102
 103                     self.items.append( channel)
 104             if not len(self.items):
 105                 log( 'OPML import finished, but no items found: %s', url, sender = self)
 106         except:
 107             log( 'Cannot import OPML from URL: %s', url, traceback=True, sender = self)
 108
 109     def format_channel( self, channel):
 110         """
 111         Formats a channel dictionary (as populated by the
 112         constructor) into a Pango markup string, suitable
 113         for output in GTK widgets.
 114
 115         The resulting string contains the title and description.
 116         """
 117         return '<b>%s</b>\n<span size="small">%s</span>' % ( xml.sax.saxutils.escape( urllib.unquote_plus( channel['title'])), xml.sax.saxutils.escape( channel['description']), )
 118
 119     def get_model( self):
 120         """
 121         Returns a gtk.ListStore with three columns:
 122
 123          - a bool that is initally set to False
 124          - a descriptive Pango markup string created
 125            by calling self.format_channel()
 126          - the URL of the channel as string
 127         """
 128         model = gtk.ListStore( gobject.TYPE_BOOLEAN, gobject.TYPE_STRING, gobject.TYPE_STRING)
 129
 130         for channel in self.items:
 131             model.append( [ False, self.format_channel( channel), channel['url'] ])
 132
 133         return model
 134
 135
 136
 137 class Exporter(object):
 138     """
 139     Helper class to export a list of channel objects
 140     to a local file in OPML 1.1 format.
 141
 142     See www.opml.org for the OPML specification.
 143     """
 144
 145     FEED_TYPE = 'rss'
 146
 147     def __init__( self, filename):
 148         if filename.endswith( '.opml') or filename.endswith( '.xml'):
 149             self.filename = filename
 150         else:
 151             self.filename = '%s.opml' % ( filename, )
 152
 153     def create_node( self, doc, name, content):
 154         """
 155         Creates a simple XML Element node in a document
 156         with tag name "name" and text content "content",
 157         as in <name>content</name> and returns the element.
 158         """
 159         node = doc.createElement( name)
 160         node.appendChild( doc.createTextNode( content))
 161         return node
 162
 163     def create_outline( self, doc, channel):
 164         """
 165         Creates a OPML outline as XML Element node in a
 166         document for the supplied channel.
 167         """
 168         outline = doc.createElement( 'outline')
 169         outline.setAttribute( 'title', channel.title)
 170         outline.setAttribute( 'text', channel.description)
 171         outline.setAttribute( 'xmlUrl', channel.url)
 172         outline.setAttribute( 'type', self.FEED_TYPE)
 173         return outline
 174
 175     def write( self, channels):
 176         """
 177         Creates a XML document containing metadata for each
 178         channel object in the "channels" parameter, which
 179         should be a list of channel objects.
 180
 181         Returns True on success or False when there was an
 182         error writing the file.
 183         """
 184         doc = xml.dom.minidom.Document()
 185
 186         opml = doc.createElement( 'opml')
 187         opml.setAttribute( 'version', '1.1')
 188         doc.appendChild( opml)
 189
 190         head = doc.createElement( 'head')
 191         head.appendChild( self.create_node( doc, 'title', 'gPodder subscriptions'))
 192         head.appendChild( self.create_node( doc, 'dateCreated', datetime.datetime.now().ctime()))
 193         opml.appendChild( head)
 194
 195         body = doc.createElement( 'body')
 196         for channel in channels:
 197             body.appendChild( self.create_outline( doc, channel))
 198         opml.appendChild( body)
 199
 200         try:
 201             data = doc.toprettyxml(encoding='utf-8', indent='    ', newl=os.linesep)
 202             # We want to have at least 512 KiB free disk space after
 203             # saving the opml data, if this is not possible, don't
 204             # try to save the new file, but keep the old one so we
 205             # don't end up with a clobbed, empty opml file.
 206             FREE_DISK_SPACE_AFTER = 1024*512
 207             if util.get_free_disk_space(os.path.dirname(self.filename)) < 2*len(data)+FREE_DISK_SPACE_AFTER:
 208                 log('Not enough free disk space to save channel list to %s', self.filename, sender = self)
 209                 return False
 210             fp = open(self.filename+'.tmp', 'w')
 211             fp.write(data)
 212             fp.close()
 213             os.rename(self.filename+'.tmp', self.filename)
 214         except:
 215             log( 'Could not open file for writing: %s', self.filename, sender = self)
 216             return False
 217
 218         return True
 219