X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Ftheplatform.py;h=bf6f82f5ae52af83706807f7112c9bdd74138461;hb=db1c969da5dc22b36228b50ac2c09f3a50dd17ae;hp=f02e0f58d5752000ab0036be78c293cbc4b4ce62;hpb=26e1c3514f4af1ed60cd1114a653fe49e1fa8d11;p=youtube-dl diff --git a/youtube_dl/extractor/theplatform.py b/youtube_dl/extractor/theplatform.py index f02e0f58d..bf6f82f5a 100644 --- a/youtube_dl/extractor/theplatform.py +++ b/youtube_dl/extractor/theplatform.py @@ -1,61 +1,64 @@ +# -*- coding: utf-8 -*- from __future__ import unicode_literals import re -import json import time import hmac import binascii import hashlib -from .common import InfoExtractor +from .once import OnceIE +from ..compat import ( + compat_parse_qs, + compat_urllib_parse_urlparse, +) from ..utils import ( - determine_ext, ExtractorError, - xpath_with_ns, - unsmuggle_url, - int_or_none, - url_basename, float_or_none, + int_or_none, + sanitized_Request, + unsmuggle_url, + xpath_with_ns, + mimetype2ext, + find_xpath_attr, ) default_ns = 'http://www.w3.org/2005/SMIL21/Language' _x = lambda p: xpath_with_ns(p, {'smil': default_ns}) -class ThePlatformBaseIE(InfoExtractor): - def _extract_theplatform_smil_formats(self, smil_url, video_id, note='Downloading SMIL data'): - meta = self._download_xml(smil_url, video_id, note=note) - try: - error_msg = next( - n.attrib['abstract'] - for n in meta.findall(_x('.//smil:ref')) - if n.attrib.get('title') == 'Geographic Restriction' or n.attrib.get('title') == 'Expired') - except StopIteration: - pass - else: - raise ExtractorError(error_msg, expected=True) +class ThePlatformBaseIE(OnceIE): + def _extract_theplatform_smil(self, smil_url, video_id, note='Downloading SMIL data'): + meta = self._download_xml(smil_url, video_id, note=note, query={'format': 'SMIL'}) + error_element = find_xpath_attr(meta, _x('.//smil:ref'), 'src') + if error_element is not None and error_element.attrib['src'].startswith( + 'http://link.theplatform.com/s/errorFiles/Unavailable.'): + raise ExtractorError(error_element.attrib['abstract'], expected=True) - formats = self._parse_smil_formats( + smil_formats = self._parse_smil_formats( meta, smil_url, video_id, namespace=default_ns, # the parameters are from syfy.com, other sites may use others, # they also work for nbc.com f4m_params={'g': 'UXWGVKRWHFSP', 'hdcore': '3.0.3'}, transform_rtmp_url=lambda streamer, src: (streamer, 'mp4:' + src)) - for _format in formats: - ext = determine_ext(_format['url']) - if ext == 'once': - _format['ext'] = 'mp4' + formats = [] + for _format in smil_formats: + if OnceIE.suitable(_format['url']): + formats.extend(self._extract_once_formats(_format['url'])) + else: + formats.append(_format) self._sort_formats(formats) - return formats + subtitles = self._parse_smil_subtitles(meta, default_ns) + + return formats, subtitles def get_metadata(self, path, video_id): info_url = 'http://link.theplatform.com/s/%s?format=preview' % path - info_json = self._download_webpage(info_url, video_id) - info = json.loads(info_json) + info = self._download_json(info_url, video_id) subtitles = {} captions = info.get('captions') @@ -63,7 +66,7 @@ class ThePlatformBaseIE(InfoExtractor): for caption in captions: lang, src, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type') subtitles[lang] = [{ - 'ext': 'srt' if mime == 'text/srt' else 'ttml', + 'ext': mimetype2ext(mime), 'url': src, }] @@ -79,7 +82,7 @@ class ThePlatformBaseIE(InfoExtractor): class ThePlatformIE(ThePlatformBaseIE): _VALID_URL = r'''(?x) (?:https?://(?:link|player)\.theplatform\.com/[sp]/(?P[^/]+)/ - (?:(?P(?:[^/]+/)+select/media/)|(?P(?:[^/\?]+/(?:swf|config)|onsite)/select/))? + (?:(?:(?:[^/]+/)+select/)?(?Pmedia/(?:guid/\d+/)?)|(?P(?:[^/\?]+/(?:swf|config)|onsite)/select/))? |theplatform:)(?P[^/\?&]+)''' _TESTS = [{ @@ -120,6 +123,24 @@ class ThePlatformIE(ThePlatformBaseIE): }, { 'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7', 'only_matching': True, + }, { + 'url': 'http://player.theplatform.com/p/2E2eJC/nbcNewsOffsite?guid=tdy_or_siri_150701', + 'md5': 'fb96bb3d85118930a5b055783a3bd992', + 'info_dict': { + 'id': 'tdy_or_siri_150701', + 'ext': 'mp4', + 'title': 'iPhone Siri’s sassy response to a math question has people talking', + 'description': 'md5:a565d1deadd5086f3331d57298ec6333', + 'duration': 83.0, + 'thumbnail': 're:^https?://.*\.jpg$', + 'timestamp': 1435752600, + 'upload_date': '20150701', + }, + }, { + # From http://www.nbc.com/the-blacklist/video/sir-crispin-crandall/2928790?onid=137781#vc137781=1 + # geo-restricted (US), HLS encrypted with AES-128 + 'url': 'http://player.theplatform.com/p/NnzsPC/onsite_universal/select/media/guid/2410887629/2928790?fwsitesection=nbc_the_blacklist_video_library&autoPlay=true&carouselID=137781', + 'only_matching': True, }] @staticmethod @@ -133,7 +154,7 @@ class ThePlatformIE(ThePlatformBaseIE): def hex_to_str(hex): return binascii.a2b_hex(hex) - relative_path = url.split('http://link.theplatform.com/s/')[1].split('?')[0] + relative_path = re.match(r'https?://link.theplatform.com/s/([^?]+)', url).group(1) clear_text = hex_to_str(flags + expiration_date + str_to_hex(relative_path)) checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest() sig = flags + expiration_date + checksum + str_to_hex(sig_secret) @@ -149,13 +170,49 @@ class ThePlatformIE(ThePlatformBaseIE): if not provider_id: provider_id = 'dJ5BDC' - path = provider_id + path = provider_id + '/' if mobj.group('media'): - path += '/media' - path += '/' + video_id + path += mobj.group('media') + path += video_id + + qs_dict = compat_parse_qs(compat_urllib_parse_urlparse(url).query) + if 'guid' in qs_dict: + webpage = self._download_webpage(url, video_id) + scripts = re.findall(r']+src="([^"]+)"', webpage) + feed_id = None + # feed id usually locates in the last script. + # Seems there's no pattern for the interested script filename, so + # I try one by one + for script in reversed(scripts): + feed_script = self._download_webpage( + self._proto_relative_url(script, 'http:'), + video_id, 'Downloading feed script') + feed_id = self._search_regex( + r'defaultFeedId\s*:\s*"([^"]+)"', feed_script, + 'default feed id', default=None) + if feed_id is not None: + break + if feed_id is None: + raise ExtractorError('Unable to find feed id') + return self.url_result('http://feed.theplatform.com/f/%s/%s?byGuid=%s' % ( + provider_id, feed_id, qs_dict['guid'][0])) if smuggled_data.get('force_smil_url', False): smil_url = url + # Explicitly specified SMIL (see https://github.com/rg3/youtube-dl/issues/7385) + elif '/guid/' in url: + headers = {} + source_url = smuggled_data.get('source_url') + if source_url: + headers['Referer'] = source_url + request = sanitized_Request(url, headers=headers) + webpage = self._download_webpage(request, video_id) + smil_url = self._search_regex( + r']+href=(["\'])(?P.+?)\1[^>]+type=["\']application/smil\+xml', + webpage, 'smil url', group='url') + path = self._search_regex( + r'link\.theplatform\.com/s/((?:[^/?#&]+/)+[^/?#&]+)', smil_url, 'path') + smil_url += '?' if '?' not in smil_url else '&' + 'formats=m3u,mpeg4' elif mobj.group('config'): config_url = url + '&form=json' config_url = config_url.replace('swf/', 'config/') @@ -165,20 +222,22 @@ class ThePlatformIE(ThePlatformBaseIE): release_url = config['releaseUrl'] else: release_url = 'http://link.theplatform.com/s/%s?mbr=true' % path - smil_url = release_url + '&format=SMIL&formats=MPEG4&manifest=f4m' + smil_url = release_url + '&formats=MPEG4&manifest=f4m' else: - smil_url = 'http://link.theplatform.com/s/%s/meta.smil?format=smil&mbr=true' % path + smil_url = 'http://link.theplatform.com/s/%s?mbr=true' % path sig = smuggled_data.get('sig') if sig: smil_url = self._sign_url(smil_url, sig['key'], sig['secret']) - formats = self._extract_theplatform_smil_formats(smil_url, video_id) + formats, subtitles = self._extract_theplatform_smil(smil_url, video_id) ret = self.get_metadata(path, video_id) + combined_subtitles = self._merge_subtitles(ret.get('subtitles', {}), subtitles) ret.update({ 'id': video_id, 'formats': formats, + 'subtitles': combined_subtitles, }) return ret @@ -190,7 +249,7 @@ class ThePlatformFeedIE(ThePlatformBaseIE): _TEST = { # From http://player.theplatform.com/p/7wvmTC/MSNBCEmbeddedOffSite?guid=n_hardball_5biden_140207 'url': 'http://feed.theplatform.com/f/7wvmTC/msnbc_video-p-test?form=json&pretty=true&range=-40&byGuid=n_hardball_5biden_140207', - 'md5': '22d2b84f058d3586efcd99e57d59d314', + 'md5': '6e32495b5073ab414471b615c5ded394', 'info_dict': { 'id': 'n_hardball_5biden_140207', 'ext': 'mp4', @@ -216,15 +275,18 @@ class ThePlatformFeedIE(ThePlatformBaseIE): entry = feed['entries'][0] formats = [] + subtitles = {} first_video_id = None duration = None for item in entry['media$content']: - smil_url = item['plfile$url'] + '&format=SMIL&Tracking=true&Embedded=true&formats=MPEG4,F4M' - cur_video_id = url_basename(smil_url) + smil_url = item['plfile$url'] + '&mbr=true' + cur_video_id = ThePlatformIE._match_id(smil_url) if first_video_id is None: first_video_id = cur_video_id duration = float_or_none(item.get('plfile$duration')) - formats.extend(self._extract_theplatform_smil_formats(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id)) + cur_formats, cur_subtitles = self._extract_theplatform_smil(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id) + formats.extend(cur_formats) + subtitles = self._merge_subtitles(subtitles, cur_subtitles) self._sort_formats(formats) @@ -238,9 +300,11 @@ class ThePlatformFeedIE(ThePlatformBaseIE): categories = [item['media$name'] for item in entry.get('media$categories', [])] ret = self.get_metadata('%s/%s' % (provider_id, first_video_id), video_id) + subtitles = self._merge_subtitles(subtitles, ret['subtitles']) ret.update({ 'id': video_id, 'formats': formats, + 'subtitles': subtitles, 'thumbnails': thumbnails, 'duration': duration, 'timestamp': timestamp,