Merge tag 'upstream/2016.08.17'

[youtubedl] / youtube_dl / extractor / twitter.py

diff --git a/youtube_dl/extractor/twitter.py b/youtube_dl/extractor/twitter.py
index 9d3e46b946843ae0da6b9de525c4aa4b8b3f4cbb..b7384298619608ab879337326b1e6719962932e3 100644 (file)
--- a/youtube_dl/extractor/twitter.py
+++ b/youtube_dl/extractor/twitter.py
@@ -4,25 +4,33 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
 import re
 
 from .common import InfoExtractor

-from ..compat import compat_urllib_request

 from ..utils import (
 from ..utils import (

+    determine_ext,

     float_or_none,
     xpath_text,
     remove_end,
     float_or_none,
     xpath_text,
     remove_end,

+    int_or_none,
+    ExtractorError,

 )
 
 
 )
 
 

-class TwitterCardIE(InfoExtractor):
+class TwitterBaseIE(InfoExtractor):
+    def _get_vmap_video_url(self, vmap_url, video_id):
+        vmap_data = self._download_xml(vmap_url, video_id)
+        return xpath_text(vmap_data, './/MediaFile').strip()
+
+
+class TwitterCardIE(TwitterBaseIE):

     IE_NAME = 'twitter:card'
     IE_NAME = 'twitter:card'

-    _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/cards/tfw/v1/(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/(?:cards/tfw/v1|videos/tweet)/(?P<id>\d+)'

     _TESTS = [
         {
             'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',
     _TESTS = [
         {
             'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',

-            'md5': '7d2f6b4d2eb841a7ccc893d479bfceb4',
+            # MD5 checksums are different in different places

             'info_dict': {
                 'id': '560070183650213889',
                 'ext': 'mp4',
             'info_dict': {
                 'id': '560070183650213889',
                 'ext': 'mp4',

-                'title': 'TwitterCard',
+                'title': 'Twitter Card',

                 'thumbnail': 're:^https?://.*\.jpg$',
                 'duration': 30.033,
             }
                 'thumbnail': 're:^https?://.*\.jpg$',
                 'duration': 30.033,
             }


@@ -33,82 +41,138 @@ class TwitterCardIE(InfoExtractor):
             'info_dict': {
                 'id': '623160978427936768',
                 'ext': 'mp4',
             'info_dict': {
                 'id': '623160978427936768',
                 'ext': 'mp4',

-                'title': 'TwitterCard',
+                'title': 'Twitter Card',

                 'thumbnail': 're:^https?://.*\.jpg',
                 'duration': 80.155,
             },
         },
         {
             'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977',
                 'thumbnail': 're:^https?://.*\.jpg',
                 'duration': 80.155,
             },
         },
         {
             'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977',

-            'md5': 'b6f35e8b08a0bec6c8af77a2f4b3a814',
+            'md5': 'd4724ffe6d2437886d004fa5de1043b3',

             'info_dict': {
                 'id': 'dq4Oj5quskI',
                 'ext': 'mp4',
                 'title': 'Ubuntu 11.10 Overview',
             'info_dict': {
                 'id': 'dq4Oj5quskI',
                 'ext': 'mp4',
                 'title': 'Ubuntu 11.10 Overview',

-                'description': 'Take a quick peek at what\'s new and improved in Ubuntu 11.10.\n\nOnce installed take a look at 10 Things to Do After Installing: http://www.omgubuntu.co.uk/2011/10/10-things-to-do-after-installing-ubuntu-11-10/',
+                'description': 'Take a quick peek at what\'s new and improved in Ubuntu 11.10.\n\nOnce installed take a look at 10 Things to Do After Installing: http://www.omgubuntu.co.uk/2011/10/10...',

                 'upload_date': '20111013',
                 'uploader': 'OMG! Ubuntu!',
                 'uploader_id': 'omgubuntu',
             },
                 'upload_date': '20111013',
                 'uploader': 'OMG! Ubuntu!',
                 'uploader_id': 'omgubuntu',
             },

-        }
+            'add_ie': ['Youtube'],
+        },
+        {
+            'url': 'https://twitter.com/i/cards/tfw/v1/665289828897005568',
+            'md5': 'ab2745d0b0ce53319a534fccaa986439',
+            'info_dict': {
+                'id': 'iBb2x00UVlv',
+                'ext': 'mp4',
+                'upload_date': '20151113',
+                'uploader_id': '1189339351084113920',
+                'uploader': 'ArsenalTerje',
+                'title': 'Vine by ArsenalTerje',
+            },
+            'add_ie': ['Vine'],
+        }, {
+            'url': 'https://twitter.com/i/videos/tweet/705235433198714880',
+            'md5': '3846d0a07109b5ab622425449b59049d',
+            'info_dict': {
+                'id': '705235433198714880',
+                'ext': 'mp4',
+                'title': 'Twitter web player',
+                'thumbnail': 're:^https?://.*\.jpg',
+            },
+        },

     ]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
     ]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 

-        # Different formats served for different User-Agents
-        USER_AGENTS = [
-            'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',  # mp4
-            'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:38.0) Gecko/20100101 Firefox/38.0',  # webm
-        ]
-

         config = None
         formats = []
         config = None
         formats = []

-        for user_agent in USER_AGENTS:
-            request = compat_urllib_request.Request(url)
-            request.add_header('User-Agent', user_agent)
-            webpage = self._download_webpage(request, video_id)
-
-            youtube_url = self._html_search_regex(
-                r'<iframe[^>]+src="((?:https?:)?//www.youtube.com/embed/[^"]+)"',
-                webpage, 'youtube iframe', default=None)
-            if youtube_url:
-                return self.url_result(youtube_url, 'Youtube')
-
-            config = self._parse_json(self._html_search_regex(
-                r'data-player-config="([^"]+)"', webpage, 'data player config'),
-                video_id)
-            if 'playlist' not in config:
-                if 'vmapUrl' in config:
-                    vmap_data = self._download_xml(config['vmapUrl'], video_id)
-                    video_url = xpath_text(vmap_data, './/MediaFile').strip()
-                    formats.append({
-                        'url': video_url,
-                    })
-                    break   # same video regardless of UA
-                continue
-
-            video_url = config['playlist'][0]['source']
-
-            f = {
-                'url': video_url,
-            }
+        duration = None
+
+        webpage = self._download_webpage(url, video_id)
+
+        iframe_url = self._html_search_regex(
+            r'<iframe[^>]+src="((?:https?:)?//(?:www.youtube.com/embed/[^"]+|(?:www\.)?vine\.co/v/\w+/card))"',
+            webpage, 'video iframe', default=None)
+        if iframe_url:
+            return self.url_result(iframe_url)

 
 

+        config = self._parse_json(self._html_search_regex(
+            r'data-(?:player-)?config="([^"]+)"', webpage, 'data player config'),
+            video_id)
+
+        if config.get('source_type') == 'vine':
+            return self.url_result(config['player_url'], 'Vine')
+
+        def _search_dimensions_in_video_url(a_format, video_url):

             m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
             if m:
             m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
             if m:

-                f.update({
+                a_format.update({

                     'width': int(m.group('width')),
                     'height': int(m.group('height')),
                 })
                     'width': int(m.group('width')),
                     'height': int(m.group('height')),
                 })

-            formats.append(f)
+
+        video_url = config.get('video_url') or config.get('playlist', [{}])[0].get('source')
+
+        if video_url:
+            if determine_ext(video_url) == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(video_url, video_id, ext='mp4', m3u8_id='hls'))
+            else:
+                f = {
+                    'url': video_url,
+                }
+
+                _search_dimensions_in_video_url(f, video_url)
+
+                formats.append(f)
+
+        vmap_url = config.get('vmapUrl') or config.get('vmap_url')
+        if vmap_url:
+            formats.append({
+                'url': self._get_vmap_video_url(vmap_url, video_id),
+            })
+
+        media_info = None
+
+        for entity in config.get('status', {}).get('entities', []):
+            if 'mediaInfo' in entity:
+                media_info = entity['mediaInfo']
+
+        if media_info:
+            for media_variant in media_info['variants']:
+                media_url = media_variant['url']
+                if media_url.endswith('.m3u8'):
+                    formats.extend(self._extract_m3u8_formats(media_url, video_id, ext='mp4', m3u8_id='hls'))
+                elif media_url.endswith('.mpd'):
+                    formats.extend(self._extract_mpd_formats(media_url, video_id, mpd_id='dash'))
+                else:
+                    vbr = int_or_none(media_variant.get('bitRate'), scale=1000)
+                    a_format = {
+                        'url': media_url,
+                        'format_id': 'http-%d' % vbr if vbr else 'http',
+                        'vbr': vbr,
+                    }
+                    # Reported bitRate may be zero
+                    if not a_format['vbr']:
+                        del a_format['vbr']
+
+                    _search_dimensions_in_video_url(a_format, media_url)
+
+                    formats.append(a_format)
+
+            duration = float_or_none(media_info.get('duration', {}).get('nanos'), scale=1e9)
+

         self._sort_formats(formats)
 
         self._sort_formats(formats)
 

-        thumbnail = config.get('posterImageUrl')
-        duration = float_or_none(config.get('duration'))
+        title = self._search_regex(r'<title>([^<]+)</title>', webpage, 'title')
+        thumbnail = config.get('posterImageUrl') or config.get('image_src')
+        duration = float_or_none(config.get('duration')) or duration

 
         return {
             'id': video_id,
 
         return {
             'id': video_id,

-            'title': 'TwitterCard',
+            'title': title,

             'thumbnail': thumbnail,
             'duration': duration,
             'formats': formats,
             'thumbnail': thumbnail,
             'duration': duration,
             'formats': formats,


@@ -120,47 +184,214 @@ class TwitterIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.|m\.|mobile\.)?twitter\.com/(?P<user_id>[^/]+)/status/(?P<id>\d+)'
     _TEMPLATE_URL = 'https://twitter.com/%s/status/%s'
 
     _VALID_URL = r'https?://(?:www\.|m\.|mobile\.)?twitter\.com/(?P<user_id>[^/]+)/status/(?P<id>\d+)'
     _TEMPLATE_URL = 'https://twitter.com/%s/status/%s'
 

-    _TEST = {
+    _TESTS = [{

         'url': 'https://twitter.com/freethenipple/status/643211948184596480',
         'url': 'https://twitter.com/freethenipple/status/643211948184596480',

-        'md5': '31cd83a116fc41f99ae3d909d4caf6a0',

         'info_dict': {
             'id': '643211948184596480',
             'ext': 'mp4',
             'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!',
             'thumbnail': 're:^https?://.*\.jpg',
         'info_dict': {
             'id': '643211948184596480',
             'ext': 'mp4',
             'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!',
             'thumbnail': 're:^https?://.*\.jpg',

-            'duration': 12.922,

             'description': 'FREE THE NIPPLE on Twitter: "FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ"',
             'uploader': 'FREE THE NIPPLE',
             'uploader_id': 'freethenipple',
         },
             'description': 'FREE THE NIPPLE on Twitter: "FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ"',
             'uploader': 'FREE THE NIPPLE',
             'uploader_id': 'freethenipple',
         },

-    }
+        'params': {
+            'skip_download': True,  # requires ffmpeg
+        },
+    }, {
+        'url': 'https://twitter.com/giphz/status/657991469417025536/photo/1',
+        'md5': 'f36dcd5fb92bf7057f155e7d927eeb42',
+        'info_dict': {
+            'id': '657991469417025536',
+            'ext': 'mp4',
+            'title': 'Gifs - tu vai cai tu vai cai tu nao eh capaz disso tu vai cai',
+            'description': 'Gifs on Twitter: "tu vai cai tu vai cai tu nao eh capaz disso tu vai cai https://t.co/tM46VHFlO5"',
+            'thumbnail': 're:^https?://.*\.png',
+            'uploader': 'Gifs',
+            'uploader_id': 'giphz',
+        },
+        'expected_warnings': ['height', 'width'],
+        'skip': 'Account suspended',
+    }, {
+        'url': 'https://twitter.com/starwars/status/665052190608723968',
+        'md5': '39b7199856dee6cd4432e72c74bc69d4',
+        'info_dict': {
+            'id': '665052190608723968',
+            'ext': 'mp4',
+            'title': 'Star Wars - A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens.',
+            'description': 'Star Wars on Twitter: "A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens."',
+            'uploader_id': 'starwars',
+            'uploader': 'Star Wars',
+        },
+    }, {
+        'url': 'https://twitter.com/BTNBrentYarina/status/705235433198714880',
+        'info_dict': {
+            'id': '705235433198714880',
+            'ext': 'mp4',
+            'title': 'Brent Yarina - Khalil Iverson\'s missed highlight dunk. And made highlight dunk. In one highlight.',
+            'description': 'Brent Yarina on Twitter: "Khalil Iverson\'s missed highlight dunk. And made highlight dunk. In one highlight."',
+            'uploader_id': 'BTNBrentYarina',
+            'uploader': 'Brent Yarina',
+        },
+        'params': {
+            # The same video as https://twitter.com/i/videos/tweet/705235433198714880
+            # Test case of TwitterCardIE
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://twitter.com/jaydingeer/status/700207533655363584',
+        'md5': '',
+        'info_dict': {
+            'id': '700207533655363584',
+            'ext': 'mp4',
+            'title': 'Donte The Dumbass - BEAT PROD: @suhmeduh #Damndaniel',
+            'description': 'Donte The Dumbass on Twitter: "BEAT PROD: @suhmeduh  https://t.co/HBrQ4AfpvZ #Damndaniel https://t.co/byBooq2ejZ"',
+            'thumbnail': 're:^https?://.*\.jpg',
+            'uploader': 'Donte The Dumbass',
+            'uploader_id': 'jaydingeer',
+        },
+        'params': {
+            'skip_download': True,  # requires ffmpeg
+        },
+    }, {
+        'url': 'https://twitter.com/Filmdrunk/status/713801302971588609',
+        'md5': '89a15ed345d13b86e9a5a5e051fa308a',
+        'info_dict': {
+            'id': 'MIOxnrUteUd',
+            'ext': 'mp4',
+            'title': 'Dr.Pepperの飲み方 #japanese #バカ #ドクペ #電動ガン',
+            'uploader': 'TAKUMA',
+            'uploader_id': '1004126642786242560',
+            'upload_date': '20140615',
+        },
+        'add_ie': ['Vine'],
+    }, {
+        'url': 'https://twitter.com/captainamerica/status/719944021058060289',
+        'info_dict': {
+            'id': '719944021058060289',
+            'ext': 'mp4',
+            'title': 'Captain America - @King0fNerd Are you sure you made the right choice? Find out in theaters.',
+            'description': 'Captain America on Twitter: "@King0fNerd Are you sure you made the right choice? Find out in theaters. https://t.co/GpgYi9xMJI"',
+            'uploader_id': 'captainamerica',
+            'uploader': 'Captain America',
+        },
+        'params': {
+            'skip_download': True,  # requires ffmpeg
+        },
+    }]

 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         user_id = mobj.group('user_id')
         twid = mobj.group('id')
 
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         user_id = mobj.group('user_id')
         twid = mobj.group('id')
 

-        webpage = self._download_webpage(self._TEMPLATE_URL % (user_id, twid), twid)
+        webpage, urlh = self._download_webpage_handle(
+            self._TEMPLATE_URL % (user_id, twid), twid)
+
+        if 'twitter.com/account/suspended' in urlh.geturl():
+            raise ExtractorError('Account suspended by Twitter.', expected=True)

 
         username = remove_end(self._og_search_title(webpage), ' on Twitter')
 
 
         username = remove_end(self._og_search_title(webpage), ' on Twitter')
 

-        title = self._og_search_description(webpage).strip('').replace('\n', ' ')
+        title = description = self._og_search_description(webpage).strip('').replace('\n', ' ').strip('“”')

 
         # strip  'https -_t.co_BJYgOjSeGA' junk from filenames
 
         # strip  'https -_t.co_BJYgOjSeGA' junk from filenames

-        mobj = re.match(r'“(.*)\s+(https?://[^ ]+)”', title)
-        title, short_url = mobj.groups()
+        title = re.sub(r'\s+(https?://[^ ]+)', '', title)

 
 

-        card_id = self._search_regex(
-            r'["\']/i/cards/tfw/v1/(\d+)', webpage, 'twitter card url')
-        card_url = 'https://twitter.com/i/cards/tfw/v1/' + card_id
-
-        return {
-            '_type': 'url_transparent',
-            'ie_key': 'TwitterCard',
+        info = {

             'uploader_id': user_id,
             'uploader': username,
             'uploader_id': user_id,
             'uploader': username,

-            'url': card_url,

             'webpage_url': url,
             'webpage_url': url,

-            'description': '%s on Twitter: "%s %s"' % (username, title, short_url),
+            'description': '%s on Twitter: "%s"' % (username, description),

             'title': username + ' - ' + title,
         }
             'title': username + ' - ' + title,
         }

+
+        mobj = re.search(r'''(?x)
+            <video[^>]+class="animated-gif"(?P<more_info>[^>]+)>\s*
+                <source[^>]+video-src="(?P<url>[^"]+)"
+        ''', webpage)
+
+        if mobj:
+            more_info = mobj.group('more_info')
+            height = int_or_none(self._search_regex(
+                r'data-height="(\d+)"', more_info, 'height', fatal=False))
+            width = int_or_none(self._search_regex(
+                r'data-width="(\d+)"', more_info, 'width', fatal=False))
+            thumbnail = self._search_regex(
+                r'poster="([^"]+)"', more_info, 'poster', fatal=False)
+            info.update({
+                'id': twid,
+                'url': mobj.group('url'),
+                'height': height,
+                'width': width,
+                'thumbnail': thumbnail,
+            })
+            return info
+
+        if 'class="PlayableMedia' in webpage:
+            info.update({
+                '_type': 'url_transparent',
+                'ie_key': 'TwitterCard',
+                'url': '%s//twitter.com/i/videos/tweet/%s' % (self.http_scheme(), twid),
+            })
+
+            return info
+
+        raise ExtractorError('There\'s no video in this tweet.')
+
+
+class TwitterAmplifyIE(TwitterBaseIE):
+    IE_NAME = 'twitter:amplify'
+    _VALID_URL = 'https?://amp\.twimg\.com/v/(?P<id>[0-9a-f\-]{36})'
+
+    _TEST = {
+        'url': 'https://amp.twimg.com/v/0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
+        'md5': '7df102d0b9fd7066b86f3159f8e81bf6',
+        'info_dict': {
+            'id': '0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
+            'ext': 'mp4',
+            'title': 'Twitter Video',
+            'thumbnail': 're:^https?://.*',
+        },
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        vmap_url = self._html_search_meta(
+            'twitter:amplify:vmap', webpage, 'vmap url')
+        video_url = self._get_vmap_video_url(vmap_url, video_id)
+
+        thumbnails = []
+        thumbnail = self._html_search_meta(
+            'twitter:image:src', webpage, 'thumbnail', fatal=False)
+
+        def _find_dimension(target):
+            w = int_or_none(self._html_search_meta(
+                'twitter:%s:width' % target, webpage, fatal=False))
+            h = int_or_none(self._html_search_meta(
+                'twitter:%s:height' % target, webpage, fatal=False))
+            return w, h
+
+        if thumbnail:
+            thumbnail_w, thumbnail_h = _find_dimension('image')
+            thumbnails.append({
+                'url': thumbnail,
+                'width': thumbnail_w,
+                'height': thumbnail_h,
+            })
+
+        video_w, video_h = _find_dimension('player')
+        formats = [{
+            'url': video_url,
+            'width': video_w,
+            'height': video_h,
+        }]
+
+        return {
+            'id': video_id,
+            'title': 'Twitter Video',
+            'formats': formats,
+            'thumbnails': thumbnails,
+        }