- return {
- 'id': track_id,
- 'title': data['title'],
- 'formats': formats,
- 'duration': float_or_none(data.get('duration')),
- }
- else:
- raise ExtractorError('No free songs found')
-
- download_link = m_download.group(1)
- video_id = self._search_regex(
- r'(?ms)var TralbumData = .*?[{,]\s*id: (?P<id>\d+),?$',
- webpage, 'video id')
-
- download_webpage = self._download_webpage(download_link, video_id, 'Downloading free downloads page')
- # We get the dictionary of the track from some javascript code
- all_info = self._parse_json(self._search_regex(
- r'(?sm)items: (.*?),$', download_webpage, 'items'), video_id)
- info = all_info[0]
- # We pick mp3-320 for now, until format selection can be easily implemented.
- mp3_info = info['downloads']['mp3-320']
- # If we try to use this url it says the link has expired
- initial_url = mp3_info['url']
- m_url = re.match(
- r'(?P<server>http://(.*?)\.bandcamp\.com)/download/track\?enc=mp3-320&fsig=(?P<fsig>.*?)&id=(?P<id>.*?)&ts=(?P<ts>.*)$',
- initial_url)
- # We build the url we will use to get the final track url
- # This url is build in Bandcamp in the script download_bunde_*.js
- request_url = '%s/statdownload/track?enc=mp3-320&fsig=%s&id=%s&ts=%s&.rand=665028774616&.vrs=1' % (m_url.group('server'), m_url.group('fsig'), video_id, m_url.group('ts'))
- final_url_webpage = self._download_webpage(request_url, video_id, 'Requesting download url')
- # If we could correctly generate the .rand field the url would be
- # in the "download_url" key
- final_url = self._proto_relative_url(self._search_regex(
- r'"retry_url":"(.+?)"', final_url_webpage, 'final video URL'), 'http:')
+ artist = extract('artist')
+ album = extract('album_title')
+ timestamp = unified_timestamp(
+ extract('publish_date') or extract('album_publish_date'))
+ release_date = unified_strdate(extract('album_release_date'))
+
+ download_link = self._search_regex(
+ r'freeDownloadPage\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1', webpage,
+ 'download link', default=None, group='url')
+ if download_link:
+ track_id = self._search_regex(
+ r'(?ms)var TralbumData = .*?[{,]\s*id: (?P<id>\d+),?$',
+ webpage, 'track id')
+
+ download_webpage = self._download_webpage(
+ download_link, track_id, 'Downloading free downloads page')
+
+ blob = self._parse_json(
+ self._search_regex(
+ r'data-blob=(["\'])(?P<blob>{.+?})\1', download_webpage,
+ 'blob', group='blob'),
+ track_id, transform_source=unescapeHTML)
+
+ info = try_get(
+ blob, (lambda x: x['digital_items'][0],
+ lambda x: x['download_items'][0]), dict)
+ if info:
+ downloads = info.get('downloads')
+ if isinstance(downloads, dict):
+ if not track:
+ track = info.get('title')
+ if not artist:
+ artist = info.get('artist')
+ if not thumbnail:
+ thumbnail = info.get('thumb_url')
+
+ download_formats = {}
+ download_formats_list = blob.get('download_formats')
+ if isinstance(download_formats_list, list):
+ for f in blob['download_formats']:
+ name, ext = f.get('name'), f.get('file_extension')
+ if all(isinstance(x, compat_str) for x in (name, ext)):
+ download_formats[name] = ext.strip('.')
+
+ for format_id, f in downloads.items():
+ format_url = f.get('url')
+ if not format_url:
+ continue
+ # Stat URL generation algorithm is reverse engineered from
+ # download_*_bundle_*.js
+ stat_url = update_url_query(
+ format_url.replace('/download/', '/statdownload/'), {
+ '.rand': int(time.time() * 1000 * random.random()),
+ })
+ format_id = f.get('encoding_name') or format_id
+ stat = self._download_json(
+ stat_url, track_id, 'Downloading %s JSON' % format_id,
+ transform_source=lambda s: s[s.index('{'):s.rindex('}') + 1],
+ fatal=False)
+ if not stat:
+ continue
+ retry_url = url_or_none(stat.get('retry_url'))
+ if not retry_url:
+ continue
+ formats.append({
+ 'url': self._proto_relative_url(retry_url, 'http:'),
+ 'ext': download_formats.get(format_id),
+ 'format_id': format_id,
+ 'format_note': f.get('description'),
+ 'filesize': parse_filesize(f.get('size_mb')),
+ 'vcodec': 'none',
+ })
+
+ self._sort_formats(formats)
+
+ title = '%s - %s' % (artist, track) if artist else track
+
+ if not duration:
+ duration = float_or_none(self._html_search_meta(
+ 'duration', webpage, default=None))
Rapsys Git / youtubedl / blobdiff