2 # -*- coding: utf-8 -*-
4 # This file is part of my.gpodder.org.
6 # my.gpodder.org is free software: you can redistribute it and/or modify it
7 # under the terms of the GNU Affero General Public License as published by
8 # the Free Software Foundation, either version 3 of the License, or (at your
9 # option) any later version.
11 # my.gpodder.org is distributed in the hope that it will be useful, but
12 # WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
13 # or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public
14 # License for more details.
16 # You should have received a copy of the GNU Affero General Public License
17 # along with my.gpodder.org. If not, see <http://www.gnu.org/licenses/>.
25 from datetime
import datetime
26 from itertools
import chain
28 from django
.conf
import settings
30 from mygpo
.core
.slugs
import assign_missing_episode_slugs
, assign_slug
, \
32 from feedservice
.parse
import parse_feed
, FetchFeedException
33 from feedservice
.parse
.text
import ConvertMarkdown
34 from feedservice
.parse
.models
import ParserException
35 from mygpo
.utils
import file_hash
, split_list
, deep_eq
36 from mygpo
.web
.logo
import CoverArt
37 from mygpo
.data
.podcast
import subscribe_at_hub
38 from mygpo
.db
.couchdb
.episode
import episode_for_podcast_id_url
, \
39 episodes_for_podcast_uncached
40 from mygpo
.db
.couchdb
.podcast
import podcast_for_url
41 from mygpo
.directory
.tags
import update_category
42 from mygpo
.decorators
import repeat_on_conflict
43 from mygpo
.db
.couchdb
import get_main_database
46 logger
= logging
.getLogger(__name__
)
49 class NoPodcastCreated(Exception):
50 """ raised when no podcast obj was created for a new URL """
53 class NoEpisodesException(Exception):
54 """ raised when parsing something that doesn't contain any episodes """
57 class PodcastUpdater(object):
58 """ Updates a number of podcasts with data from their feeds """
61 """ Queue is an iterable of podcast objects """
62 self
.db
= get_main_database()
65 def update_queue(self
, queue
):
66 """ Fetch data for the URLs supplied as the queue iterable """
68 for n
, podcast_url
in enumerate(queue
):
69 logger
.info('Update %d - %s', n
, podcast_url
)
71 yield self
.update(podcast_url
)
73 except NoPodcastCreated
as npc
:
74 logger
.info('No podcast created: %s', npc
)
77 def update(self
, podcast_url
):
78 """ Update the podcast for the supplied URL """
81 parsed
= self
._fetch
_feed
(podcast_url
)
82 self
._validate
_parsed
(parsed
)
84 except (ParserException
, FetchFeedException
, NoEpisodesException
) as ex
:
86 # if we fail to parse the URL, we don't even create the
88 p
= podcast_for_url(podcast_url
, create
=False)
90 # if it exists already, we mark it as outdated
91 self
._mark
_outdated
(p
)
95 raise NoPodcastCreated(ex
)
97 assert parsed
, 'fetch_feed must return something'
98 p
= podcast_for_url(podcast_url
, create
=True)
99 self
._update
_podcast
(p
, parsed
)
103 def verify_podcast_url(self
, podcast_url
):
104 parsed
= self
._fetch
_feed
(podcast_url
)
105 self
._validate
_parsed
(parsed
)
109 def _fetch_feed(self
, podcast_url
):
110 return parse_feed(podcast_url
, text_processor
=ConvertMarkdown())
114 def _validate_parsed(self
, parsed
):
115 """ validates the parsed results and raises an exception if invalid
117 feedparser parses pretty much everything. We reject anything that
118 doesn't look like a feed"""
120 if not parsed
or not parsed
.episodes
:
121 raise NoEpisodesException('no episodes found')
124 @repeat_on_conflict(['podcast'])
125 def _update_podcast(self
, podcast
, parsed
):
126 """ updates a podcast according to new parser results """
128 # we need that later to decide if we can "bump" a category
129 prev_latest_episode_timestamp
= podcast
.latest_episode_timestamp
131 old_json
= copy
.deepcopy(podcast
.to_json())
133 podcast
.title
= parsed
.title
or podcast
.title
134 podcast
.urls
= list(set(podcast
.urls
+ parsed
.urls
))
135 podcast
.description
= parsed
.description
or podcast
.description
136 podcast
.subtitle
= parsed
.subtitle
or podcast
.subtitle
137 podcast
.link
= parsed
.link
or podcast
.link
138 podcast
.logo_url
= parsed
.logo
or podcast
.logo_url
139 podcast
.author
= parsed
.author
or podcast
.author
140 podcast
.language
= parsed
.language
or podcast
.language
141 podcast
.content_types
= parsed
.content_types
or podcast
.content_types
142 podcast
.tags
['feed'] = parsed
.tags
or podcast
.tags
.get('feed', [])
143 podcast
.common_episode_title
= parsed
.common_title
or podcast
.common_episode_title
144 podcast
.new_location
= parsed
.new_location
or podcast
.new_location
145 podcast
.flattr_url
= parsed
.flattr
or podcast
.flattr_url
146 podcast
.hub
= parsed
.hub
or podcast
.hub
147 podcast
.license
= parsed
.license
or podcast
.license
150 if podcast
.new_location
:
151 new_podcast
= podcast_for_url(podcast
.new_location
)
152 if new_podcast
!= podcast
:
153 self
._mark
_outdated
(podcast
, 'redirected to different podcast')
156 elif not new_podcast
:
157 podcast
.urls
.insert(0, podcast
.new_location
)
160 episodes
= self
._update
_episodes
(podcast
, parsed
.episodes
)
162 # latest episode timestamp
163 eps
= filter(lambda e
: bool(e
.released
), episodes
)
164 eps
= sorted(eps
, key
=lambda e
: e
.released
)
166 podcast
.latest_episode_timestamp
= eps
[-1].released
167 podcast
.episode_count
= len(eps
)
170 self
._update
_categories
(podcast
, prev_latest_episode_timestamp
)
172 # try to download the logo and reset logo_url to None on http errors
173 found
= self
._save
_podcast
_logo
(podcast
.logo_url
)
175 podcast
.logo_url
= None
177 if not deep_eq(old_json
, podcast
.to_json()):
178 logger
.info('Saving podcast.')
179 podcast
.last_update
= datetime
.utcnow()
183 subscribe_at_hub(podcast
)
185 assign_slug(podcast
, PodcastSlug
)
186 assign_missing_episode_slugs(podcast
)
189 def _update_categories(self
, podcast
, prev_timestamp
):
190 """ checks some practical requirements and updates a category """
192 from datetime
import timedelta
194 max_timestamp
= datetime
.utcnow() + timedelta(days
=1)
197 if not podcast
.latest_episode_timestamp
:
201 if prev_timestamp
and podcast
.latest_episode_timestamp
<= prev_timestamp
:
204 # too far in the future
205 if podcast
.latest_episode_timestamp
> max_timestamp
:
208 # not enough subscribers
209 if podcast
.subscriber_count() < settings
.MIN_SUBSCRIBERS_CATEGORY
:
212 update_category(podcast
)
215 @repeat_on_conflict(['podcast'])
216 def _update_episodes(self
, podcast
, parsed_episodes
):
218 all_episodes
= set(episodes_for_podcast_uncached(podcast
))
219 remaining
= list(all_episodes
)
220 updated_episodes
= []
222 for parsed_episode
in parsed_episodes
:
226 for f
in parsed_episode
.files
:
233 guid
= parsed_episode
.guid
235 # pop matchin episodes out of the "existing" list
236 matching
, remaining
= split_list(remaining
, lambda e
: (e
.guid
and e
.guid
== guid
) or url
in e
.urls
)
239 new_episode
= episode_for_podcast_id_url(podcast
.get_id(),
241 matching
= [new_episode
]
242 all_episodes
.add(new_episode
)
245 for episode
in matching
:
246 old_json
= copy
.deepcopy(episode
.to_json())
248 episode
.guid
= parsed_episode
.guid
or episode
.guid
249 episode
.title
= parsed_episode
.title
or episode
.title
250 episode
.description
= parsed_episode
.description
or episode
.description
251 episode
.subtitle
= parsed_episode
.subtitle
or episode
.subtitle
252 episode
.content
= parsed_episode
.content
or parsed_episode
.description
or episode
.content
253 episode
.link
= parsed_episode
.link
or episode
.link
254 episode
.released
= datetime
.utcfromtimestamp(parsed_episode
.released
) if parsed_episode
.released
else episode
.released
255 episode
.author
= parsed_episode
.author
or episode
.author
256 episode
.duration
= parsed_episode
.duration
or episode
.duration
257 episode
.filesize
= parsed_episode
.files
[0].filesize
258 episode
.language
= parsed_episode
.language
or episode
.language
259 episode
.mimetypes
= list(set(filter(None, [f
.mimetype
for f
in parsed_episode
.files
])))
260 episode
.flattr_url
= parsed_episode
.flattr
or episode
.flattr_url
261 episode
.license
= parsed_episode
.license
or episode
.license
263 urls
= list(chain
.from_iterable(f
.urls
for f
in parsed_episode
.files
))
264 episode
.urls
= sorted(set(episode
.urls
+ urls
), key
=len)
266 if not deep_eq(old_json
, episode
.to_json()):
267 episode
.last_update
= datetime
.utcnow()
268 updated_episodes
.append(episode
)
271 outdated_episodes
= all_episodes
- set(updated_episodes
)
273 # set episodes to be outdated, where necessary
274 for e
in filter(lambda e
: not e
.outdated
, outdated_episodes
):
276 updated_episodes
.append(e
)
280 logger
.info('Updating %d episodes', len(updated_episodes
))
281 self
.db
.save_docs(updated_episodes
)
286 def _save_podcast_logo(self
, cover_art
):
291 image_sha1
= hashlib
.sha1(cover_art
).hexdigest()
292 prefix
= CoverArt
.get_prefix(image_sha1
)
294 filename
= CoverArt
.get_original(prefix
, image_sha1
)
295 dirname
= CoverArt
.get_dir(filename
)
297 # get hash of existing file
298 if os
.path
.exists(filename
):
299 with
open(filename
) as f
:
300 old_hash
= file_hash(f
).digest()
304 logger
.info('Logo %s', cover_art
)
307 with
open(filename
, 'w') as fp
:
308 fp
.write(urllib2
.urlopen(cover_art
).read())
310 # get hash of new file
311 with
open(filename
) as f
:
312 new_hash
= file_hash(f
).digest()
314 # remove thumbnails if cover changed
315 if old_hash
!= new_hash
:
316 thumbnails
= CoverArt
.get_existing_thumbnails(prefix
, filename
)
317 logger
.info('Removing %d thumbnails', len(thumbnails
))
323 except (urllib2
.HTTPError
, urllib2
.URLError
, ValueError,
324 httplib
.BadStatusLine
) as e
:
325 logger
.warn('Exception while updating podcast: %s', str(e
))
328 @repeat_on_conflict(['podcast'])
329 def _mark_outdated(self
, podcast
, msg
=''):
330 logger
.info('marking podcast outdated: %s', msg
)
331 podcast
.outdated
= True
332 podcast
.last_update
= datetime
.utcnow()
334 self
._update
_episodes
(podcast
, [])