[funnyordie] move extraction to VoxMedia extractor and improve vox volume embed extraction(closes #16846)

youtube_dl/extractor/extractors.py

@@ -396,7 +396,6 @@ from .frontendmasters import (
 )
 from .funimation import FunimationIE
 from .funk import FunkIE
-from .funnyordie import FunnyOrDieIE
 from .fusion import FusionIE
 from .fxnetworks import FXNetworksIE
 from .gaia import GaiaIE

youtube_dl/extractor/funnyordie.py

@@ -1,162 +0,0 @@
-from __future__ import unicode_literals
-
-import re
-
-from .common import InfoExtractor
-from ..utils import (
-    ExtractorError,
-    float_or_none,
-    int_or_none,
-    unified_timestamp,
-)
-
-
-class FunnyOrDieIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?funnyordie\.com/(?P<type>embed|articles|videos)/(?P<id>[0-9a-f]+)(?:$|[?#/])'
-    _TESTS = [{
-        'url': 'http://www.funnyordie.com/videos/0732f586d7/heart-shaped-box-literal-video-version',
-        'md5': 'bcd81e0c4f26189ee09be362ad6e6ba9',
-        'info_dict': {
-            'id': '0732f586d7',
-            'ext': 'mp4',
-            'title': 'Heart-Shaped Box: Literal Video Version',
-            'description': 'md5:ea09a01bc9a1c46d9ab696c01747c338',
-            'thumbnail': r're:^http:.*\.jpg$',
-            'uploader': 'DASjr',
-            'timestamp': 1317904928,
-            'upload_date': '20111006',
-            'duration': 318.3,
-        },
-    }, {
-        'url': 'http://www.funnyordie.com/embed/e402820827',
-        'info_dict': {
-            'id': 'e402820827',
-            'ext': 'mp4',
-            'title': 'Please Use This Song (Jon Lajoie)',
-            'description': 'Please use this to sell something.  www.jonlajoie.com',
-            'thumbnail': r're:^http:.*\.jpg$',
-            'timestamp': 1398988800,
-            'upload_date': '20140502',
-        },
-        'params': {
-            'skip_download': True,
-        },
-    }, {
-        'url': 'http://www.funnyordie.com/articles/ebf5e34fc8/10-hours-of-walking-in-nyc-as-a-man',
-        'only_matching': True,
-    }]
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-
-        video_id = mobj.group('id')
-        webpage = self._download_webpage(url, video_id)
-
-        links = re.findall(r'<source src="([^"]+/v)[^"]+\.([^"]+)" type=\'video', webpage)
-        if not links:
-            raise ExtractorError('No media links available for %s' % video_id)
-
-        links.sort(key=lambda link: 1 if link[1] == 'mp4' else 0)
-
-        m3u8_url = self._search_regex(
-            r'<source[^>]+src=(["\'])(?P<url>.+?/master\.m3u8[^"\']*)\1',
-            webpage, 'm3u8 url', group='url')
-
-        formats = []
-
-        m3u8_formats = self._extract_m3u8_formats(
-            m3u8_url, video_id, 'mp4', 'm3u8_native',
-            m3u8_id='hls', fatal=False)
-        source_formats = list(filter(
-            lambda f: f.get('vcodec') != 'none', m3u8_formats))
-
-        bitrates = [int(bitrate) for bitrate in re.findall(r'[,/]v(\d+)(?=[,/])', m3u8_url)]
-        bitrates.sort()
-
-        if source_formats:
-            self._sort_formats(source_formats)
-
-        for bitrate, f in zip(bitrates, source_formats or [{}] * len(bitrates)):
-            for path, ext in links:
-                ff = f.copy()
-                if ff:
-                    if ext != 'mp4':
-                        ff = dict(
-                            [(k, v) for k, v in ff.items()
-                             if k in ('height', 'width', 'format_id')])
-                    ff.update({
-                        'format_id': ff['format_id'].replace('hls', ext),
-                        'ext': ext,
-                        'protocol': 'http',
-                    })
-                else:
-                    ff.update({
-                        'format_id': '%s-%d' % (ext, bitrate),
-                        'vbr': bitrate,
-                    })
-                ff['url'] = self._proto_relative_url(
-                    '%s%d.%s' % (path, bitrate, ext))
-                formats.append(ff)
-        self._check_formats(formats, video_id)
-
-        formats.extend(m3u8_formats)
-        self._sort_formats(
-            formats, field_preference=('height', 'width', 'tbr', 'format_id'))
-
-        subtitles = {}
-        for src, src_lang in re.findall(r'<track kind="captions" src="([^"]+)" srclang="([^"]+)"', webpage):
-            subtitles[src_lang] = [{
-                'ext': src.split('/')[-1],
-                'url': 'http://www.funnyordie.com%s' % src,
-            }]
-
-        timestamp = unified_timestamp(self._html_search_meta(
-            'uploadDate', webpage, 'timestamp', default=None))
-
-        uploader = self._html_search_regex(
-            r'<h\d[^>]+\bclass=["\']channel-preview-name[^>]+>(.+?)</h',
-            webpage, 'uploader', default=None)
-
-        title, description, thumbnail, duration = [None] * 4
-
-        medium = self._parse_json(
-            self._search_regex(
-                r'jsonMedium\s*=\s*({.+?});', webpage, 'JSON medium',
-                default='{}'),
-            video_id, fatal=False)
-        if medium:
-            title = medium.get('title')
-            duration = float_or_none(medium.get('duration'))
-            if not timestamp:
-                timestamp = unified_timestamp(medium.get('publishDate'))
-
-        post = self._parse_json(
-            self._search_regex(
-                r'fb_post\s*=\s*(\{.*?\});', webpage, 'post details',
-                default='{}'),
-            video_id, fatal=False)
-        if post:
-            if not title:
-                title = post.get('name')
-            description = post.get('description')
-            thumbnail = post.get('picture')
-
-        if not title:
-            title = self._og_search_title(webpage)
-        if not description:
-            description = self._og_search_description(webpage)
-        if not duration:
-            duration = int_or_none(self._html_search_meta(
-                ('video:duration', 'duration'), webpage, 'duration', default=False))
-
-        return {
-            'id': video_id,
-            'title': title,
-            'description': description,
-            'thumbnail': thumbnail,
-            'uploader': uploader,
-            'timestamp': timestamp,
-            'duration': duration,
-            'formats': formats,
-            'subtitles': subtitles,
-        }

youtube_dl/extractor/voxmedia.py

@@ -4,7 +4,10 @@ from __future__ import unicode_literals
 from .common import InfoExtractor
 from .once import OnceIE
 from ..compat import compat_urllib_parse_unquote
-from ..utils import ExtractorError
+from ..utils import (
+    ExtractorError,
+    int_or_none,
+)
 
 
 class VoxMediaVolumeIE(OnceIE):
@@ -13,18 +16,43 @@ class VoxMediaVolumeIE(OnceIE):
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
-        video_data = self._parse_json(self._search_regex(
+
-            r'Volume\.createVideo\(({.+})\s*,\s*{.*}\s*,\s*\[.*\]\s*,\s*{.*}\);', webpage, 'video data'), video_id)
+        setup = self._parse_json(self._search_regex(
-        for provider_video_type in ('ooyala', 'youtube', 'brightcove'):
+            r'setup\s*=\s*({.+});', webpage, 'setup'), video_id)
-            provider_video_id = video_data.get('%s_id' % provider_video_type)
+        video_data = setup.get('video') or {}
-            if not provider_video_id:
-                continue
         info = {
             'id': video_id,
             'title': video_data.get('title_short'),
             'description': video_data.get('description_long') or video_data.get('description_short'),
             'thumbnail': video_data.get('brightcove_thumbnail')
         }
+        asset = setup.get('asset') or setup.get('params') or {}
+
+        formats = []
+        hls_url = asset.get('hls_url')
+        if hls_url:
+            formats.extend(self._extract_m3u8_formats(
+                hls_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
+        mp4_url = asset.get('mp4_url')
+        if mp4_url:
+            tbr = self._search_regex(r'-(\d+)k\.', mp4_url, 'bitrate', default=None)
+            format_id = 'http'
+            if tbr:
+                format_id += '-' + tbr
+            formats.append({
+                'format_id': format_id,
+                'url': mp4_url,
+                'tbr': int_or_none(tbr),
+            })
+        if formats:
+            self._sort_formats(formats)
+            info['formats'] = formats
+            return info
+
+        for provider_video_type in ('ooyala', 'youtube', 'brightcove'):
+            provider_video_id = video_data.get('%s_id' % provider_video_type)
+            if not provider_video_id:
+                continue
             if provider_video_type == 'brightcove':
                 info['formats'] = self._extract_once_formats(provider_video_id)
                 self._sort_formats(info['formats'])
@@ -39,46 +67,49 @@ class VoxMediaVolumeIE(OnceIE):
 
 
 class VoxMediaIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?(?:(?:theverge|vox|sbnation|eater|polygon|curbed|racked)\.com|recode\.net)/(?:[^/]+/)*(?P<id>[^/?]+)'
+    _VALID_URL = r'https?://(?:www\.)?(?:(?:theverge|vox|sbnation|eater|polygon|curbed|racked|funnyordie)\.com|recode\.net)/(?:[^/]+/)*(?P<id>[^/?]+)'
     _TESTS = [{
+        # Volume embed, Youtube
         'url': 'http://www.theverge.com/2014/6/27/5849272/material-world-how-google-discovered-what-software-is-made-of',
         'info_dict': {
-            'id': '11eXZobjrG8DCSTgrNjVinU-YmmdYjhe',
+            'id': 'j4mLW6x17VM',
             'ext': 'mp4',
-            'title': 'Google\'s new material design direction',
+            'title': 'Material world: how Google discovered what software is made of',
-            'description': 'md5:2f44f74c4d14a1f800ea73e1c6832ad2',
+            'description': 'md5:dfc17e7715e3b542d66e33a109861382',
+            'upload_date': '20190710',
+            'uploader_id': 'TheVerge',
+            'uploader': 'The Verge',
         },
-        'params': {
+        'add_ie': ['Youtube'],
-            # m3u8 download
-            'skip_download': True,
-        },
-        'add_ie': ['Ooyala'],
     }, {
-        # data-ooyala-id
+        # Volume embed, Youtube
         'url': 'http://www.theverge.com/2014/10/21/7025853/google-nexus-6-hands-on-photos-video-android-phablet',
-        'md5': 'd744484ff127884cd2ba09e3fa604e4b',
+        'md5': '4c8f4a0937752b437c3ebc0ed24802b5',
         'info_dict': {
-            'id': 'RkZXU4cTphOCPDMZg5oEounJyoFI0g-B',
+            'id': 'Gy8Md3Eky38',
             'ext': 'mp4',
             'title': 'The Nexus 6: hands-on with Google\'s phablet',
-            'description': 'md5:87a51fe95ff8cea8b5bdb9ac7ae6a6af',
+            'description': 'md5:d9f0216e5fb932dd2033d6db37ac3f1d',
+            'uploader_id': 'TheVerge',
+            'upload_date': '20141021',
+            'uploader': 'The Verge',
         },
-        'add_ie': ['Ooyala'],
+        'add_ie': ['Youtube'],
-        'skip': 'Video Not Found',
+        'skip': 'similar to the previous test',
     }, {
-        # volume embed
+        # Volume embed, Youtube
         'url': 'http://www.vox.com/2016/3/31/11336640/mississippi-lgbt-religious-freedom-bill',
         'info_dict': {
-            'id': 'wydzk3dDpmRz7PQoXRsTIX6XTkPjYL0b',
+            'id': 'YCjDnX-Xzhg',
             'ext': 'mp4',
-            'title': 'The new frontier of LGBTQ civil rights, explained',
+            'title': "Mississippi's laws are so bad that its anti-LGBTQ law isn't needed to allow discrimination",
-            'description': 'md5:0dc58e94a465cbe91d02950f770eb93f',
+            'description': 'md5:fc1317922057de31cd74bce91eb1c66c',
+            'uploader_id': 'voxdotcom',
+            'upload_date': '20150915',
+            'uploader': 'Vox',
         },
-        'params': {
+        'add_ie': ['Youtube'],
-            # m3u8 download
+        'skip': 'similar to the previous test',
-            'skip_download': True,
-        },
-        'add_ie': ['Ooyala'],
     }, {
         # youtube embed
         'url': 'http://www.vox.com/2016/3/24/11291692/robot-dance',
@@ -93,6 +124,7 @@ class VoxMediaIE(InfoExtractor):
             'uploader': 'Vox',
         },
         'add_ie': ['Youtube'],
+        'skip': 'Page no longer contain videos',
     }, {
         # SBN.VideoLinkset.entryGroup multiple ooyala embeds
         'url': 'http://www.sbnation.com/college-football-recruiting/2015/2/3/7970291/national-signing-day-rationalizations-itll-be-ok-itll-be-ok',
@@ -118,10 +150,11 @@ class VoxMediaIE(InfoExtractor):
                 'description': 'md5:e02d56b026d51aa32c010676765a690d',
             },
         }],
+        'skip': 'Page no longer contain videos',
     }, {
         # volume embed, Brightcove Once
         'url': 'https://www.recode.net/2014/6/17/11628066/post-post-pc-ceo-the-full-code-conference-video-of-microsofts-satya',
-        'md5': '01571a896281f77dc06e084138987ea2',
+        'md5': '2dbc77b8b0bff1894c2fce16eded637d',
         'info_dict': {
             'id': '1231c973d',
             'ext': 'mp4',