X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fbandcamp.py;h=f14b407dc82cf8f945f581ab059bd6218866eb57;hb=HEAD;hp=dcbbdef4346c36c789e49531df1dc602bc35255b;hpb=fa19dfccf9ced09657bc27c870ba75a19690ebb7;p=youtube-dl diff --git a/youtube_dl/extractor/bandcamp.py b/youtube_dl/extractor/bandcamp.py index dcbbdef43..f14b407dc 100644 --- a/youtube_dl/extractor/bandcamp.py +++ b/youtube_dl/extractor/bandcamp.py @@ -1,142 +1,417 @@ from __future__ import unicode_literals -import json +import random import re +import time from .common import InfoExtractor -from ..utils import ( +from ..compat import ( compat_str, compat_urlparse, +) +from ..utils import ( ExtractorError, + float_or_none, + int_or_none, + KNOWN_EXTENSIONS, + parse_filesize, + str_or_none, + try_get, + unescapeHTML, + update_url_query, + unified_strdate, + unified_timestamp, + url_or_none, ) 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', - 'file': '1812978515.mp3', 'md5': 'c557841d5e50261777a6585648adf439', 'info_dict': { - "title": "youtube-dl \"'/\\\u00e4\u21ad - youtube-dl test song \"'/\\\u00e4\u21ad", - "duration": 9.8485, + 'id': '1812978515', + 'ext': 'mp3', + 'title': "youtube-dl \"'/\\\u00e4\u21ad - youtube-dl test song \"'/\\\u00e4\u21ad", + 'duration': 9.8485, }, '_skip': 'There is a limit of 200 free downloads / month for the test song' + }, { + # free download + 'url': 'http://benprunty.bandcamp.com/track/lanius-battle', + 'md5': '853e35bf34aa1d6fe2615ae612564b36', + 'info_dict': { + 'id': '2650410135', + 'ext': 'aiff', + 'title': 'Ben Prunty - Lanius (Battle)', + 'thumbnail': r're:^https?://.*\.jpg$', + 'uploader': 'Ben Prunty', + 'timestamp': 1396508491, + 'upload_date': '20140403', + 'release_date': '20140403', + 'duration': 260.877, + 'track': 'Lanius (Battle)', + 'track_number': 1, + 'track_id': '2650410135', + 'artist': 'Ben Prunty', + 'album': 'FTL: Advanced Edition Soundtrack', + }, + }, { + # no free download, mp3 128 + 'url': 'https://relapsealumni.bandcamp.com/track/hail-to-fire', + 'md5': 'fec12ff55e804bb7f7ebeb77a800c8b7', + 'info_dict': { + 'id': '2584466013', + 'ext': 'mp3', + 'title': 'Mastodon - Hail to Fire', + 'thumbnail': r're:^https?://.*\.jpg$', + 'uploader': 'Mastodon', + 'timestamp': 1322005399, + 'upload_date': '20111122', + 'release_date': '20040207', + 'duration': 120.79, + 'track': 'Hail to Fire', + 'track_number': 5, + 'track_id': '2584466013', + 'artist': 'Mastodon', + 'album': 'Call of the Mastodon', + }, }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) title = mobj.group('title') webpage = self._download_webpage(url, title) - m_download = re.search(r'freeDownloadPage: "(.*?)"', webpage) - if not m_download: - m_trackinfo = re.search(r'trackinfo: (.+),\s*?\n', webpage) - if m_trackinfo: - json_code = m_trackinfo.group(1) - data = json.loads(json_code)[0] - - formats = [] - for format_id, format_url in data['file'].items(): + thumbnail = self._html_search_meta('og:image', webpage, default=None) + + track_id = None + track = None + track_number = None + duration = None + + formats = [] + track_info = self._parse_json( + self._search_regex( + r'trackinfo\s*:\s*\[\s*({.+?})\s*\]\s*,\s*?\n', + webpage, 'track info', default='{}'), title) + if track_info: + file_ = track_info.get('file') + if isinstance(file_, dict): + for format_id, format_url in file_.items(): + if not url_or_none(format_url): + continue ext, abr_str = format_id.split('-', 1) formats.append({ 'format_id': format_id, - 'url': format_url, + 'url': self._proto_relative_url(format_url, 'http:'), 'ext': ext, 'vcodec': 'none', 'acodec': ext, - 'abr': int(abr_str), + 'abr': int_or_none(abr_str), }) + track = track_info.get('title') + track_id = str_or_none(track_info.get('track_id') or track_info.get('id')) + track_number = int_or_none(track_info.get('track_num')) + duration = float_or_none(track_info.get('duration')) - self._sort_formats(formats) + def extract(key): + return self._search_regex( + r'\b%s\s*["\']?\s*:\s*(["\'])(?P<value>(?:(?!\1).)+)\1' % key, + webpage, key, default=None, group='value') - return { - 'id': compat_str(data['id']), - 'title': data['title'], - 'formats': formats, - 'duration': float(data['duration']), - } - else: - raise ExtractorError('No free songs found') - - download_link = m_download.group(1) - video_id = re.search( - r'var TralbumData = {(.*?)id: (?P<id>\d*?)$', - webpage, re.MULTILINE | re.DOTALL).group('id') - - download_webpage = self._download_webpage(download_link, video_id, 'Downloading free downloads page') - # We get the dictionary of the track from some javascript code - info = re.search(r'items: (.*?),$', download_webpage, re.MULTILINE).group(1) - info = json.loads(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'] - re_url = r'(?P<server>http://(.*?)\.bandcamp\.com)/download/track\?enc=mp3-320&fsig=(?P<fsig>.*?)&id=(?P<id>.*?)&ts=(?P<ts>.*)$' - m_url = re.match(re_url, initial_url) - #We build the url we will use to get the final track url - # This url is build in Bandcamp in the script download_bunde_*.js - request_url = '%s/statdownload/track?enc=mp3-320&fsig=%s&id=%s&ts=%s&.rand=665028774616&.vrs=1' % (m_url.group('server'), m_url.group('fsig'), video_id, m_url.group('ts')) - final_url_webpage = self._download_webpage(request_url, video_id, 'Requesting download url') - # If we could correctly generate the .rand field the url would be - #in the "download_url" key - final_url = re.search(r'"retry_url":"(.*?)"', final_url_webpage).group(1) + artist = extract('artist') + album = extract('album_title') + timestamp = unified_timestamp( + extract('publish_date') or extract('album_publish_date')) + release_date = unified_strdate(extract('album_release_date')) + + download_link = self._search_regex( + r'freeDownloadPage\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1', webpage, + 'download link', default=None, group='url') + if download_link: + track_id = self._search_regex( + r'(?ms)var TralbumData = .*?[{,]\s*id: (?P<id>\d+),?$', + 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)) return { - 'id': video_id, - 'title': info['title'], - 'ext': 'mp3', - 'vcodec': 'none', - 'url': final_url, - 'thumbnail': info.get('thumb_url'), - 'uploader': info.get('artist'), + 'id': track_id, + 'title': title, + 'thumbnail': thumbnail, + 'uploader': artist, + 'timestamp': timestamp, + 'release_date': release_date, + 'duration': duration, + 'track': track, + 'track_number': track_number, + 'track_id': track_id, + 'artist': artist, + 'album': album, + 'formats': formats, } class BandcampAlbumIE(InfoExtractor): IE_NAME = 'Bandcamp:album' - _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<title>[^?#]+))' + _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^/?#&]+))?' - _TEST = { + _TESTS = [{ 'url': 'http://blazo.bandcamp.com/album/jazz-format-mixtape-vol-1', 'playlist': [ { - 'file': '1353101989.mp3', 'md5': '39bc1eded3476e927c724321ddf116cf', 'info_dict': { + 'id': '1353101989', + 'ext': 'mp3', 'title': 'Intro', } }, { - 'file': '38097443.mp3', 'md5': '1a2c32e2691474643e912cc6cd4bffaa', 'info_dict': { + 'id': '38097443', + 'ext': 'mp3', 'title': 'Kero One - Keep It Alive (Blazo remix)', } }, ], + 'info_dict': { + 'title': 'Jazz Format Mixtape vol.1', + 'id': 'jazz-format-mixtape-vol-1', + 'uploader_id': 'blazo', + }, 'params': { 'playlistend': 2 }, - 'skip': 'Bandcamp imposes download limits. See test_playlists:test_bandcamp_album for the playlist test' - } + 'skip': 'Bandcamp imposes download limits.' + }, { + 'url': 'http://nightbringer.bandcamp.com/album/hierophany-of-the-open-grave', + 'info_dict': { + 'title': 'Hierophany of the Open Grave', + 'uploader_id': 'nightbringer', + 'id': 'hierophany-of-the-open-grave', + }, + 'playlist_mincount': 9, + }, { + 'url': 'http://dotscale.bandcamp.com', + 'info_dict': { + 'title': 'Loom', + 'id': 'dotscale', + 'uploader_id': 'dotscale', + }, + 'playlist_mincount': 7, + }, { + # with escaped quote in title + 'url': 'https://jstrecords.bandcamp.com/album/entropy-ep', + 'info_dict': { + 'title': '"Entropy" EP', + 'uploader_id': 'jstrecords', + '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) - playlist_id = mobj.group('subdomain') - title = mobj.group('title') - display_id = title or playlist_id - webpage = self._download_webpage(url, display_id) - tracks_paths = re.findall(r'<a href="(.*?)" itemprop="url">', webpage) - if not tracks_paths: + 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) + 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] - title = self._search_regex(r'album_title : "(.*?)"', webpage, 'title') + 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) + if title: + title = title.replace(r'\"', '"') return { '_type': 'playlist', + 'uploader_id': uploader_id, 'id': playlist_id, - 'display_id': display_id, '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 url_or_none(format_url): + 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 + }