X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fpbs.py;h=f11d5da5271d91fb15f9ee006743b1360897b08e;hb=54fc90aabfb71968f28af68dfe3f7a3544cc2f0b;hp=744e4a09aa6facf850aeda790bebf687c90a14a4;hpb=7ba71e30fb9ec188d122f79a914f688928e80e91;p=youtube-dl diff --git a/youtube_dl/extractor/pbs.py b/youtube_dl/extractor/pbs.py index 744e4a09a..f11d5da52 100644 --- a/youtube_dl/extractor/pbs.py +++ b/youtube_dl/extractor/pbs.py @@ -8,7 +8,11 @@ from ..utils import ( ExtractorError, determine_ext, int_or_none, + float_or_none, + js_to_json, + orderedSet, strip_jsonp, + strip_or_none, unified_strdate, US_RATINGS, ) @@ -16,7 +20,7 @@ from ..utils import ( class PBSIE(InfoExtractor): _STATIONS = ( - (r'(?:video|www)\.pbs\.org', 'PBS: Public Broadcasting Service'), # http://www.pbs.org/ + (r'(?:video|www|player)\.pbs\.org', 'PBS: Public Broadcasting Service'), # http://www.pbs.org/ (r'video\.aptv\.org', 'APT - Alabama Public Television (WBIQ)'), # http://aptv.org/ (r'video\.gpb\.org', 'GPB/Georgia Public Broadcasting (WGTV)'), # http://www.gpb.org/ (r'video\.mpbonline\.org', 'Mississippi Public Broadcasting (WMPN)'), # http://www.mpbonline.org @@ -183,42 +187,38 @@ class PBSIE(InfoExtractor): _VALID_URL = r'''(?x)https?:// (?: # Direct video URL - (?:%s)/(?:viralplayer|video)/(?P[0-9]+)/? | + (?:%s)/(?:(?:vir|port)alplayer|video)/(?P[0-9]+)(?:[?/]|$) | # Article with embedded player (or direct video) - (?:www\.)?pbs\.org/(?:[^/]+/){2,5}(?P[^/]+?)(?:\.html)?/?(?:$|[?\#]) | + (?:www\.)?pbs\.org/(?:[^/]+/){1,5}(?P[^/]+?)(?:\.html)?/?(?:$|[?\#]) | # Player (?:video|player)\.pbs\.org/(?:widget/)?partnerplayer/(?P[^/]+)/ ) ''' % '|'.join(list(zip(*_STATIONS))[0]) + _GEO_COUNTRIES = ['US'] + _TESTS = [ { 'url': 'http://www.pbs.org/tpt/constitution-usa-peter-sagal/watch/a-more-perfect-union/', - 'md5': 'ce1888486f0908d555a8093cac9a7362', + 'md5': '173dc391afd361fa72eab5d3d918968d', 'info_dict': { 'id': '2365006249', 'ext': 'mp4', 'title': 'Constitution USA with Peter Sagal - A More Perfect Union', - 'description': 'md5:ba0c207295339c8d6eced00b7c363c6a', + 'description': 'md5:31b664af3c65fd07fa460d306b837d00', 'duration': 3190, }, - 'params': { - 'skip_download': True, # requires ffmpeg - }, }, { 'url': 'http://www.pbs.org/wgbh/pages/frontline/losing-iraq/', - 'md5': '143c98aa54a346738a3d78f54c925321', + 'md5': '6f722cb3c3982186d34b0f13374499c7', 'info_dict': { 'id': '2365297690', 'ext': 'mp4', 'title': 'FRONTLINE - Losing Iraq', - 'description': 'md5:f5bfbefadf421e8bb8647602011caf8e', + 'description': 'md5:5979a4d069b157f622d02bff62fbe654', 'duration': 5050, }, - 'params': { - 'skip_download': True, # requires ffmpeg - } }, { 'url': 'http://www.pbs.org/newshour/bb/education-jan-june12-cyberschools_02-23/', @@ -227,7 +227,7 @@ class PBSIE(InfoExtractor): 'id': '2201174722', 'ext': 'mp4', 'title': 'PBS NewsHour - Cyber Schools Gain Popularity, but Quality Questions Persist', - 'description': 'md5:5871c15cba347c1b3d28ac47a73c7c28', + 'description': 'md5:86ab9a3d04458b876147b355788b8781', 'duration': 801, }, }, @@ -237,13 +237,10 @@ class PBSIE(InfoExtractor): 'info_dict': { 'id': '2365297708', 'ext': 'mp4', - 'description': 'md5:68d87ef760660eb564455eb30ca464fe', 'title': 'Great Performances - Dudamel Conducts Verdi Requiem at the Hollywood Bowl - Full', + 'description': 'md5:657897370e09e2bc6bf0f8d2cd313c6b', 'duration': 6559, - 'thumbnail': 're:^https?://.*\.jpg$', - }, - 'params': { - 'skip_download': True, # requires ffmpeg + 'thumbnail': r're:^https?://.*\.jpg$', }, }, { @@ -256,13 +253,10 @@ class PBSIE(InfoExtractor): 'description': 'md5:c741d14e979fc53228c575894094f157', 'title': 'NOVA - Killer Typhoon', 'duration': 3172, - 'thumbnail': 're:^https?://.*\.jpg$', + 'thumbnail': r're:^https?://.*\.jpg$', 'upload_date': '20140122', 'age_limit': 10, }, - 'params': { - 'skip_download': True, # requires ffmpeg - }, }, { 'url': 'http://www.pbs.org/wgbh/pages/frontline/united-states-of-secrets/', @@ -271,6 +265,13 @@ class PBSIE(InfoExtractor): }, 'playlist_count': 2, }, + { + 'url': 'http://www.pbs.org/wgbh/americanexperience/films/great-war/', + 'info_dict': { + 'id': 'great-war', + }, + 'playlist_count': 3, + }, { 'url': 'http://www.pbs.org/wgbh/americanexperience/films/death/player/', 'info_dict': { @@ -278,46 +279,41 @@ class PBSIE(InfoExtractor): 'display_id': 'player', 'ext': 'mp4', 'title': 'American Experience - Death and the Civil War, Chapter 1', - 'description': 'American Experience, TV’s most-watched history series, brings to life the compelling stories from our past that inform our understanding of the world today.', + 'description': 'md5:67fa89a9402e2ee7d08f53b920674c18', 'duration': 682, - 'thumbnail': 're:^https?://.*\.jpg$', + 'thumbnail': r're:^https?://.*\.jpg$', }, 'params': { 'skip_download': True, # requires ffmpeg }, }, { - 'url': 'http://video.pbs.org/video/2365367186/', + 'url': 'http://www.pbs.org/video/2365245528/', + 'md5': '115223d41bd55cda8ae5cd5ed4e11497', 'info_dict': { - 'id': '2365367186', - 'display_id': '2365367186', + 'id': '2365245528', + 'display_id': '2365245528', 'ext': 'mp4', - 'title': 'To Catch A Comet - Full Episode', - 'description': 'On November 12, 2014, billions of kilometers from Earth, spacecraft orbiter Rosetta and lander Philae did what no other had dared to attempt \u2014 land on the volatile surface of a comet as it zooms around the sun at 67,000 km/hr. The European Space Agency hopes this mission can help peer into our past and unlock secrets of our origins.', - 'duration': 3342, - 'thumbnail': 're:^https?://.*\.jpg$', + 'title': 'FRONTLINE - United States of Secrets (Part One)', + 'description': 'md5:55756bd5c551519cc4b7703e373e217e', + 'duration': 6851, + 'thumbnail': r're:^https?://.*\.jpg$', }, - 'params': { - 'skip_download': True, # requires ffmpeg - }, - 'skip': 'Expired', }, { # Video embedded in iframe containing angle brackets as attribute's value (e.g. # "', webpage): url = self._search_regex( @@ -406,6 +453,9 @@ class PBSIE(InfoExtractor): if url: break + if not url: + url = self._og_search_url(webpage) + mobj = re.match(self._VALID_URL, url) player_id = mobj.group('player_id') @@ -421,10 +471,10 @@ class PBSIE(InfoExtractor): video_id = mobj.group('id') display_id = video_id - return video_id, display_id, None + return video_id, display_id, None, description def _real_extract(self, url): - video_id, display_id, upload_date = self._extract_webpage(url) + video_id, display_id, upload_date, description = self._extract_webpage(url) if isinstance(video_id, list): entries = [self.url_result( @@ -432,29 +482,69 @@ class PBSIE(InfoExtractor): for vid_id in video_id] return self.playlist_result(entries, display_id) - info = self._download_json( - 'http://player.pbs.org/videoInfo/%s?format=json&type=partner' % video_id, - display_id) + info = None + redirects = [] + redirect_urls = set() + + def extract_redirect_urls(info): + for encoding_name in ('recommended_encoding', 'alternate_encoding'): + redirect = info.get(encoding_name) + if not redirect: + continue + redirect_url = redirect.get('url') + if redirect_url and redirect_url not in redirect_urls: + redirects.append(redirect) + redirect_urls.add(redirect_url) + + chapters = [] + # Player pages may also serve different qualities + for page in ('widget/partnerplayer', 'portalplayer'): + player = self._download_webpage( + 'http://player.pbs.org/%s/%s' % (page, video_id), + display_id, 'Downloading %s page' % page, fatal=False) + if player: + video_info = self._parse_json( + self._search_regex( + r'(?s)PBS\.videoData\s*=\s*({.+?});\n', + player, '%s video data' % page, default='{}'), + display_id, transform_source=js_to_json, fatal=False) + if video_info: + extract_redirect_urls(video_info) + if not info: + info = video_info + if not chapters: + for chapter_data in re.findall(r'(?s)chapters\.push\(({.*?})\)', player): + chapter = self._parse_json(chapter_data, video_id, js_to_json, fatal=False) + if not chapter: + continue + start_time = float_or_none(chapter.get('start_time'), 1000) + duration = float_or_none(chapter.get('duration'), 1000) + if start_time is None or duration is None: + continue + chapters.append({ + 'start_time': start_time, + 'end_time': start_time + duration, + 'title': chapter.get('title'), + }) formats = [] - for encoding_name in ('recommended_encoding', 'alternate_encoding'): - redirect = info.get(encoding_name) - if not redirect: - continue - redirect_url = redirect.get('url') - if not redirect_url: - continue + http_url = None + for num, redirect in enumerate(redirects): + redirect_id = redirect.get('eeid') redirect_info = self._download_json( - redirect_url + '?format=json', display_id, - 'Downloading %s video url info' % encoding_name) + '%s?format=json' % redirect['url'], display_id, + 'Downloading %s video url info' % (redirect_id or num), + headers=self.geo_verification_headers()) if redirect_info['status'] == 'error': + message = self._ERRORS.get( + redirect_info['http_code'], redirect_info['message']) + if redirect_info['http_code'] == 403: + self.raise_geo_restricted( + msg=message, countries=self._GEO_COUNTRIES) raise ExtractorError( - '%s said: %s' % ( - self.IE_NAME, - self._ERRORS.get(redirect_info['http_code'], redirect_info['message'])), - expected=True) + '%s said: %s' % (self.IE_NAME, message), expected=True) format_url = redirect_info.get('url') if not format_url: @@ -462,12 +552,41 @@ class PBSIE(InfoExtractor): if determine_ext(format_url) == 'm3u8': formats.extend(self._extract_m3u8_formats( - format_url, display_id, 'mp4', preference=1, m3u8_id='hls')) + format_url, display_id, 'mp4', m3u8_id='hls', fatal=False)) else: formats.append({ 'url': format_url, - 'format_id': redirect.get('eeid'), + 'format_id': redirect_id, }) + if re.search(r'^https?://.*(?:\d+k|baseline)', format_url): + http_url = format_url + self._remove_duplicate_formats(formats) + m3u8_formats = list(filter( + lambda f: f.get('protocol') == 'm3u8' and f.get('vcodec') != 'none', + formats)) + if http_url: + for m3u8_format in m3u8_formats: + bitrate = self._search_regex(r'(\d+)k', m3u8_format['url'], 'bitrate', default=None) + # Lower qualities (150k and 192k) are not available as HTTP formats (see [1]), + # we won't try extracting them. + # Since summer 2016 higher quality formats (4500k and 6500k) are also available + # albeit they are not documented in [2]. + # 1. https://github.com/rg3/youtube-dl/commit/cbc032c8b70a038a69259378c92b4ba97b42d491#commitcomment-17313656 + # 2. https://projects.pbs.org/confluence/display/coveapi/COVE+Video+Specifications + if not bitrate or int(bitrate) < 400: + continue + f_url = re.sub(r'\d+k|baseline', bitrate + 'k', http_url) + # This may produce invalid links sometimes (e.g. + # http://www.pbs.org/wgbh/frontline/film/suicide-plan) + if not self._is_valid_url(f_url, display_id, 'http-%sk video' % bitrate): + continue + f = m3u8_format.copy() + f.update({ + 'url': f_url, + 'format_id': m3u8_format['format_id'].replace('hls', 'http'), + 'protocol': 'http', + }) + formats.append(f) self._sort_formats(formats) rating_str = info.get('rating') @@ -482,22 +601,39 @@ class PBSIE(InfoExtractor): 'ext': 'ttml', 'url': closed_captions_url, }] + mobj = re.search(r'/(\d+)_Encoded\.dfxp', closed_captions_url) + if mobj: + ttml_caption_suffix, ttml_caption_id = mobj.group(0, 1) + ttml_caption_id = int(ttml_caption_id) + subtitles['en'].extend([{ + 'url': closed_captions_url.replace( + ttml_caption_suffix, '/%d_Encoded.srt' % (ttml_caption_id + 1)), + 'ext': 'srt', + }, { + 'url': closed_captions_url.replace( + ttml_caption_suffix, '/%d_Encoded.vtt' % (ttml_caption_id + 2)), + 'ext': 'vtt', + }]) # info['title'] is often incomplete (e.g. 'Full Episode', 'Episode 5', etc) # Try turning it to 'program - title' naming scheme if possible alt_title = info.get('program', {}).get('title') if alt_title: - info['title'] = alt_title + ' - ' + re.sub(r'^' + alt_title + '[\s\-:]+', '', info['title']) + info['title'] = alt_title + ' - ' + re.sub(r'^' + alt_title + r'[\s\-:]+', '', info['title']) + + description = info.get('description') or info.get( + 'program', {}).get('description') or description return { 'id': video_id, 'display_id': display_id, 'title': info['title'], - 'description': info['program'].get('description'), + 'description': description, 'thumbnail': info.get('image_url'), 'duration': int_or_none(info.get('duration')), 'age_limit': age_limit, 'upload_date': upload_date, 'formats': formats, 'subtitles': subtitles, + 'chapters': chapters, }