X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/47d80ec0b18245caeb97018d4c1af18d0b5b972b..0390e4903cb42a95d11dfc2457d921eb811c606d:/youtube_dl/extractor/bandcamp.py?ds=sidebyside diff --git a/youtube_dl/extractor/bandcamp.py b/youtube_dl/extractor/bandcamp.py index 88c590e..be41bd5 100644 --- a/youtube_dl/extractor/bandcamp.py +++ b/youtube_dl/extractor/bandcamp.py @@ -14,14 +14,16 @@ from ..utils import ( ExtractorError, float_or_none, int_or_none, + KNOWN_EXTENSIONS, parse_filesize, unescapeHTML, update_url_query, + unified_strdate, ) class BandcampIE(InfoExtractor): - _VALID_URL = r'https?://.*?\.bandcamp\.com/track/(?P.*)' + _VALID_URL = r'https?://.*?\.bandcamp\.com/track/(?P<title>[^/?#&]+)' _TESTS = [{ 'url': 'http://youtube-dl.bandcamp.com/track/youtube-dl-test-song', 'md5': 'c557841d5e50261777a6585648adf439', @@ -34,12 +36,12 @@ class BandcampIE(InfoExtractor): '_skip': 'There is a limit of 200 free downloads / month for the test song' }, { 'url': 'http://benprunty.bandcamp.com/track/lanius-battle', - 'md5': '73d0b3171568232574e45652f8720b5c', + 'md5': '0369ace6b939f0927e62c67a1a8d9fa7', 'info_dict': { 'id': '2650410135', - 'ext': 'mp3', - 'title': 'Lanius (Battle)', - 'uploader': 'Ben Prunty Music', + 'ext': 'aiff', + 'title': 'Ben Prunty - Lanius (Battle)', + 'uploader': 'Ben Prunty', }, }] @@ -47,6 +49,7 @@ class BandcampIE(InfoExtractor): mobj = re.match(self._VALID_URL, url) title = mobj.group('title') webpage = self._download_webpage(url, title) + thumbnail = self._html_search_meta('og:image', webpage, default=None) m_download = re.search(r'freeDownloadPage: "(.*?)"', webpage) if not m_download: m_trackinfo = re.search(r'trackinfo: (.+),\s*?\n', webpage) @@ -75,6 +78,7 @@ class BandcampIE(InfoExtractor): return { 'id': track_id, 'title': data['title'], + 'thumbnail': thumbnail, 'formats': formats, 'duration': float_or_none(data.get('duration')), } @@ -143,7 +147,7 @@ class BandcampIE(InfoExtractor): return { 'id': video_id, 'title': title, - 'thumbnail': info.get('thumb_url'), + 'thumbnail': info.get('thumb_url') or thumbnail, 'uploader': info.get('artist'), 'artist': artist, 'track': track, @@ -153,7 +157,7 @@ class BandcampIE(InfoExtractor): class BandcampAlbumIE(InfoExtractor): IE_NAME = 'Bandcamp:album' - _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^?#]+)|/?(?:$|[?#]))' + _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^/?#&]+))?' _TESTS = [{ 'url': 'http://blazo.bandcamp.com/album/jazz-format-mixtape-vol-1', @@ -209,20 +213,44 @@ class BandcampAlbumIE(InfoExtractor): 'id': 'entropy-ep', }, 'playlist_mincount': 3, + }, { + # not all tracks have songs + 'url': 'https://insulters.bandcamp.com/album/we-are-the-plague', + 'info_dict': { + 'id': 'we-are-the-plague', + 'title': 'WE ARE THE PLAGUE', + 'uploader_id': 'insulters', + }, + 'playlist_count': 2, }] + @classmethod + def suitable(cls, url): + return (False + if BandcampWeeklyIE.suitable(url) or BandcampIE.suitable(url) + else super(BandcampAlbumIE, cls).suitable(url)) + def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) uploader_id = mobj.group('subdomain') album_id = mobj.group('album_id') playlist_id = album_id or uploader_id webpage = self._download_webpage(url, playlist_id) - tracks_paths = re.findall(r'<a href="(.*?)" itemprop="url">', webpage) - if not tracks_paths: + track_elements = re.findall( + r'(?s)<div[^>]*>(.*?<a[^>]+href="([^"]+?)"[^>]+itemprop="url"[^>]*>.*?)</div>', webpage) + if not track_elements: raise ExtractorError('The page doesn\'t contain any tracks') + # Only tracks with duration info have songs entries = [ - self.url_result(compat_urlparse.urljoin(url, t_path), ie=BandcampIE.ie_key()) - for t_path in tracks_paths] + self.url_result( + compat_urlparse.urljoin(url, t_path), + ie=BandcampIE.ie_key(), + video_title=self._search_regex( + r'<span\b[^>]+\bitemprop=["\']name["\'][^>]*>([^<]+)', + elem_content, 'track title', fatal=False)) + for elem_content, t_path in track_elements + if self._html_search_meta('duration', elem_content, default=None)] + title = self._html_search_regex( r'album_title\s*:\s*"((?:\\.|[^"\\])+?)"', webpage, 'title', fatal=False) @@ -235,3 +263,92 @@ class BandcampAlbumIE(InfoExtractor): 'title': title, 'entries': entries, } + + +class BandcampWeeklyIE(InfoExtractor): + IE_NAME = 'Bandcamp:weekly' + _VALID_URL = r'https?://(?:www\.)?bandcamp\.com/?\?(?:.*?&)?show=(?P<id>\d+)' + _TESTS = [{ + 'url': 'https://bandcamp.com/?show=224', + 'md5': 'b00df799c733cf7e0c567ed187dea0fd', + 'info_dict': { + 'id': '224', + 'ext': 'opus', + 'title': 'BC Weekly April 4th 2017 - Magic Moments', + 'description': 'md5:5d48150916e8e02d030623a48512c874', + 'duration': 5829.77, + 'release_date': '20170404', + 'series': 'Bandcamp Weekly', + 'episode': 'Magic Moments', + 'episode_number': 208, + 'episode_id': '224', + } + }, { + 'url': 'https://bandcamp.com/?blah/blah@&show=228', + 'only_matching': True + }] + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + + blob = self._parse_json( + self._search_regex( + r'data-blob=(["\'])(?P<blob>{.+?})\1', webpage, + 'blob', group='blob'), + video_id, transform_source=unescapeHTML) + + show = blob['bcw_show'] + + # This is desired because any invalid show id redirects to `bandcamp.com` + # which happens to expose the latest Bandcamp Weekly episode. + show_id = int_or_none(show.get('show_id')) or int_or_none(video_id) + + formats = [] + for format_id, format_url in show['audio_stream'].items(): + if not isinstance(format_url, compat_str): + continue + for known_ext in KNOWN_EXTENSIONS: + if known_ext in format_id: + ext = known_ext + break + else: + ext = None + formats.append({ + 'format_id': format_id, + 'url': format_url, + 'ext': ext, + 'vcodec': 'none', + }) + self._sort_formats(formats) + + title = show.get('audio_title') or 'Bandcamp Weekly' + subtitle = show.get('subtitle') + if subtitle: + title += ' - %s' % subtitle + + episode_number = None + seq = blob.get('bcw_seq') + + if seq and isinstance(seq, list): + try: + episode_number = next( + int_or_none(e.get('episode_number')) + for e in seq + if isinstance(e, dict) and int_or_none(e.get('id')) == show_id) + except StopIteration: + pass + + return { + 'id': video_id, + 'title': title, + 'description': show.get('desc') or show.get('short_desc'), + 'duration': float_or_none(show.get('audio_duration')), + 'is_live': False, + 'release_date': unified_strdate(show.get('published_date')), + 'series': 'Bandcamp Weekly', + 'episode': show.get('subtitle'), + 'episode_number': episode_number, + 'episode_id': compat_str(video_id), + 'formats': formats + }