debian/README.source: Change 'markup'.

[youtubedl] / youtube_dl / extractor / ted.py

diff --git a/youtube_dl/extractor/ted.py b/youtube_dl/extractor/ted.py
index 8550380779168a80b95e526f8921059e2eddf8f4..db5a4f44e6b9be254729256fc741d6f3eb3693f7 100644 (file)
--- a/youtube_dl/extractor/ted.py
+++ b/youtube_dl/extractor/ted.py
@@ -3,19 +3,28 @@ from __future__ import unicode_literals
 import json
 import re
 
 import json
 import re
 

-from .subtitles import SubtitlesInfoExtractor
+from .common import InfoExtractor

 
 

-from ..utils import (
+from ..compat import (

     compat_str,
     compat_str,

+    compat_urlparse
+)
+from ..utils import (
+    extract_attributes,
+    float_or_none,
+    int_or_none,
+    try_get,
+    url_or_none,

 )
 
 
 )
 
 

-class TEDIE(SubtitlesInfoExtractor):
+class TEDIE(InfoExtractor):
+    IE_NAME = 'ted'

     _VALID_URL = r'''(?x)
         (?P<proto>https?://)
     _VALID_URL = r'''(?x)
         (?P<proto>https?://)

-        (?P<type>www|embed)(?P<urlmain>\.ted\.com/
+        (?P<type>www|embed(?:-ssl)?)(?P<urlmain>\.ted\.com/

         (
         (

-            (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
+            (?P<type_playlist>playlists(?:/(?P<playlist_id>\d+))?) # We have a playlist

             |
             ((?P<type_talk>talks)) # We have a simple talk
             |
             |
             ((?P<type_talk>talks)) # We have a simple talk
             |


@@ -27,31 +36,42 @@ class TEDIE(SubtitlesInfoExtractor):
         '''
     _TESTS = [{
         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
         '''
     _TESTS = [{
         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',

-        'md5': 'fc94ac279feebbce69f21c0c6ee82810',
+        'md5': 'b0ce2b05ca215042124fbc9e3886493a',

         'info_dict': {
             'id': '102',
             'ext': 'mp4',
             'title': 'The illusion of consciousness',
             'description': ('Philosopher Dan Dennett makes a compelling '
         'info_dict': {
             'id': '102',
             'ext': 'mp4',
             'title': 'The illusion of consciousness',
             'description': ('Philosopher Dan Dennett makes a compelling '

-                'argument that not only don\'t we understand our own '
-                'consciousness, but that half the time our brains are '
-                'actively fooling us.'),
+                            'argument that not only don\'t we understand our own '
+                            'consciousness, but that half the time our brains are '
+                            'actively fooling us.'),

             'uploader': 'Dan Dennett',
             'uploader': 'Dan Dennett',

-            'width': 854,
+            'width': 853,

             'duration': 1308,
             'duration': 1308,

-        }
+            'view_count': int,
+            'comment_count': int,
+            'tags': list,
+        },
+        'params': {
+            'skip_download': True,
+        },

     }, {
     }, {

-        'url': 'http://www.ted.com/watch/ted-institute/ted-bcg/vishal-sikka-the-beauty-and-power-of-algorithms',
-        'md5': '226f4fb9c62380d11b7995efa4c87994',
+        # missing HTTP bitrates
+        'url': 'https://www.ted.com/talks/vishal_sikka_the_beauty_and_power_of_algorithms',

         'info_dict': {
         'info_dict': {

-            'id': 'vishal-sikka-the-beauty-and-power-of-algorithms',
+            'id': '6069',

             'ext': 'mp4',
             'ext': 'mp4',

-            'title': 'Vishal Sikka: The beauty and power of algorithms',
-            'thumbnail': 're:^https?://.+\.jpg',
-            'description': 'Adaptive, intelligent, and consistent, algorithms are emerging as the ultimate app for everything from matching consumers to products to assessing medical diagnoses. Vishal Sikka shares his appreciation for the algorithm, charting both its inherent beauty and its growing power.',
-        }
+            'title': 'The beauty and power of algorithms',
+            'thumbnail': r're:^https?://.+\.jpg',
+            'description': 'md5:734e352710fb00d840ab87ae31aaf688',
+            'uploader': 'Vishal Sikka',
+        },
+        'params': {
+            'skip_download': True,
+        },

     }, {
         'url': 'http://www.ted.com/talks/gabby_giffords_and_mark_kelly_be_passionate_be_courageous_be_your_best',
     }, {
         'url': 'http://www.ted.com/talks/gabby_giffords_and_mark_kelly_be_passionate_be_courageous_be_your_best',

+        'md5': 'e6b9617c01a7970ceac8bb2c92c346c0',

         'info_dict': {
             'id': '1972',
             'ext': 'mp4',
         'info_dict': {
             'id': '1972',
             'ext': 'mp4',


@@ -60,11 +80,15 @@ class TEDIE(SubtitlesInfoExtractor):
             'description': 'md5:5174aed4d0f16021b704120360f72b92',
             'duration': 1128,
         },
             'description': 'md5:5174aed4d0f16021b704120360f72b92',
             'duration': 1128,
         },

+        'params': {
+            'skip_download': True,
+        },

     }, {
         'url': 'http://www.ted.com/playlists/who_are_the_hackers',
         'info_dict': {
             'id': '10',
             'title': 'Who are the hackers?',
     }, {
         'url': 'http://www.ted.com/playlists/who_are_the_hackers',
         'info_dict': {
             'id': '10',
             'title': 'Who are the hackers?',

+            'description': 'md5:49a0dbe8fb76d81a0e64b4a80af7f15a'

         },
         'playlist_mincount': 6,
     }, {
         },
         'playlist_mincount': 6,
     }, {


@@ -73,7 +97,7 @@ class TEDIE(SubtitlesInfoExtractor):
         'add_ie': ['Youtube'],
         'info_dict': {
             'id': '_ZG8HBuDjgc',
         'add_ie': ['Youtube'],
         'info_dict': {
             'id': '_ZG8HBuDjgc',

-            'ext': 'mp4',
+            'ext': 'webm',

             'title': 'Douglas Adams: Parrots the Universe and Everything',
             'description': 'md5:01ad1e199c49ac640cb1196c0e9016af',
             'uploader': 'University of California Television (UCTV)',
             'title': 'Douglas Adams: Parrots the Universe and Everything',
             'description': 'md5:01ad1e199c49ac640cb1196c0e9016af',
             'uploader': 'University of California Television (UCTV)',


@@ -83,22 +107,39 @@ class TEDIE(SubtitlesInfoExtractor):
         'params': {
             'skip_download': True,
         },
         'params': {
             'skip_download': True,
         },

+    }, {
+        # no nativeDownloads
+        'url': 'https://www.ted.com/talks/tom_thum_the_orchestra_in_my_mouth',
+        'info_dict': {
+            'id': '1792',
+            'ext': 'mp4',
+            'title': 'The orchestra in my mouth',
+            'description': 'md5:5d1d78650e2f8dfcbb8ebee2951ac29a',
+            'uploader': 'Tom Thum',
+            'view_count': int,
+            'comment_count': int,
+            'tags': list,
+        },
+        'params': {
+            'skip_download': True,
+        },

     }]
 
     _NATIVE_FORMATS = {
     }]
 
     _NATIVE_FORMATS = {

-        'low': {'preference': 1, 'width': 320, 'height': 180},
-        'medium': {'preference': 2, 'width': 512, 'height': 288},
-        'high': {'preference': 3, 'width': 854, 'height': 480},
+        'low': {'width': 320, 'height': 180},
+        'medium': {'width': 512, 'height': 288},
+        'high': {'width': 854, 'height': 480},

     }
 
     def _extract_info(self, webpage):
     }
 
     def _extract_info(self, webpage):

-        info_json = self._search_regex(r'q\("\w+.init",({.+})\)</script>',
+        info_json = self._search_regex(
+            r'(?s)q\(\s*"\w+.init"\s*,\s*({.+?})\)\s*</script>',

             webpage, 'info json')
         return json.loads(info_json)
 
     def _real_extract(self, url):
         m = re.match(self._VALID_URL, url, re.VERBOSE)
             webpage, 'info json')
         return json.loads(info_json)
 
     def _real_extract(self, url):
         m = re.match(self._VALID_URL, url, re.VERBOSE)

-        if m.group('type') == 'embed':
+        if m.group('type').startswith('embed'):

             desktop_url = m.group('proto') + 'www' + m.group('urlmain')
             return self.url_result(desktop_url, 'TED')
         name = m.group('name')
             desktop_url = m.group('proto') + 'www' + m.group('urlmain')
             return self.url_result(desktop_url, 'TED')
         name = m.group('name')


@@ -113,94 +154,182 @@ class TEDIE(SubtitlesInfoExtractor):
         '''Returns the videos of the playlist'''
 
         webpage = self._download_webpage(url, name,
         '''Returns the videos of the playlist'''
 
         webpage = self._download_webpage(url, name,

-            'Downloading playlist webpage')
-        info = self._extract_info(webpage)
-        playlist_info = info['playlist']
+                                         'Downloading playlist webpage')
+
+        playlist_entries = []
+        for entry in re.findall(r'(?s)<[^>]+data-ga-context=["\']playlist["\'][^>]*>', webpage):
+            attrs = extract_attributes(entry)
+            entry_url = compat_urlparse.urljoin(url, attrs['href'])
+            playlist_entries.append(self.url_result(entry_url, self.ie_key()))
+
+        final_url = self._og_search_url(webpage, fatal=False)
+        playlist_id = (
+            re.match(self._VALID_URL, final_url).group('playlist_id')
+            if final_url else None)

 
 

-        playlist_entries = [
-            self.url_result('http://www.ted.com/talks/' + talk['slug'], self.ie_key())
-            for talk in info['talks']
-        ]

         return self.playlist_result(
         return self.playlist_result(

-            playlist_entries,
-            playlist_id=compat_str(playlist_info['id']),
-            playlist_title=playlist_info['title'])
+            playlist_entries, playlist_id=playlist_id,
+            playlist_title=self._og_search_title(webpage, fatal=False),
+            playlist_description=self._og_search_description(webpage))

 
     def _talk_info(self, url, video_name):
         webpage = self._download_webpage(url, video_name)
 
     def _talk_info(self, url, video_name):
         webpage = self._download_webpage(url, video_name)

-        self.report_extraction(video_name)

 
 

-        talk_info = self._extract_info(webpage)['talks'][0]
+        info = self._extract_info(webpage)

 
 

-        if talk_info.get('external') is not None:
-            self.to_screen('Found video from %s' % talk_info['external']['service'])
-            return {
-                '_type': 'url',
-                'url': talk_info['external']['uri'],
-            }
+        data = try_get(info, lambda x: x['__INITIAL_DATA__'], dict) or info
+        talk_info = data['talks'][0]
+
+        title = talk_info['title'].strip()
+
+        native_downloads = try_get(
+            talk_info,
+            (lambda x: x['downloads']['nativeDownloads'],
+             lambda x: x['nativeDownloads']),
+            dict) or {}

 
         formats = [{
             'url': format_url,
             'format_id': format_id,
             'format': format_id,
 
         formats = [{
             'url': format_url,
             'format_id': format_id,
             'format': format_id,

-        } for (format_id, format_url) in talk_info['nativeDownloads'].items() if format_url is not None]
+        } for (format_id, format_url) in native_downloads.items() if format_url is not None]

         if formats:
             for f in formats:
                 finfo = self._NATIVE_FORMATS.get(f['format_id'])
                 if finfo:
                     f.update(finfo)
         if formats:
             for f in formats:
                 finfo = self._NATIVE_FORMATS.get(f['format_id'])
                 if finfo:
                     f.update(finfo)

-        else:
-            # Use rtmp downloads
-            formats = [{
-                'format_id': f['name'],
-                'url': talk_info['streamer'],
-                'play_path': f['file'],
-                'ext': 'flv',
-                'width': f['width'],
-                'height': f['height'],
-                'tbr': f['bitrate'],
-            } for f in talk_info['resources']['rtmp']]
+
+        player_talk = talk_info['player_talks'][0]
+
+        external = player_talk.get('external')
+        if isinstance(external, dict):
+            service = external.get('service')
+            if isinstance(service, compat_str):
+                ext_url = None
+                if service.lower() == 'youtube':
+                    ext_url = external.get('code')
+
+                return self.url_result(ext_url or external['uri'])
+
+        resources_ = player_talk.get('resources') or talk_info.get('resources')
+
+        http_url = None
+        for format_id, resources in resources_.items():
+            if format_id == 'h264':
+                for resource in resources:
+                    h264_url = resource.get('file')
+                    if not h264_url:
+                        continue
+                    bitrate = int_or_none(resource.get('bitrate'))
+                    formats.append({
+                        'url': h264_url,
+                        'format_id': '%s-%sk' % (format_id, bitrate),
+                        'tbr': bitrate,
+                    })
+                    if re.search(r'\d+k', h264_url):
+                        http_url = h264_url
+            elif format_id == 'rtmp':
+                streamer = talk_info.get('streamer')
+                if not streamer:
+                    continue
+                for resource in resources:
+                    formats.append({
+                        'format_id': '%s-%s' % (format_id, resource.get('name')),
+                        'url': streamer,
+                        'play_path': resource['file'],
+                        'ext': 'flv',
+                        'width': int_or_none(resource.get('width')),
+                        'height': int_or_none(resource.get('height')),
+                        'tbr': int_or_none(resource.get('bitrate')),
+                    })
+            elif format_id == 'hls':
+                if not isinstance(resources, dict):
+                    continue
+                stream_url = url_or_none(resources.get('stream'))
+                if not stream_url:
+                    continue
+                formats.extend(self._extract_m3u8_formats(
+                    stream_url, video_name, 'mp4', m3u8_id=format_id,
+                    fatal=False))
+
+        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)
+                if not bitrate:
+                    continue
+                bitrate_url = re.sub(r'\d+k', bitrate, http_url)
+                if not self._is_valid_url(
+                        bitrate_url, video_name, '%s bitrate' % bitrate):
+                    continue
+                f = m3u8_format.copy()
+                f.update({
+                    'url': bitrate_url,
+                    'format_id': m3u8_format['format_id'].replace('hls', 'http'),
+                    'protocol': 'http',
+                })
+                if f.get('acodec') == 'none':
+                    del f['acodec']
+                formats.append(f)
+
+        audio_download = talk_info.get('audioDownload')
+        if audio_download:
+            formats.append({
+                'url': audio_download,
+                'format_id': 'audio',
+                'vcodec': 'none',
+            })
+

         self._sort_formats(formats)
 
         video_id = compat_str(talk_info['id'])
         self._sort_formats(formats)
 
         video_id = compat_str(talk_info['id'])

-        # subtitles
-        video_subtitles = self.extract_subtitles(video_id, talk_info)
-        if self._downloader.params.get('listsubtitles', False):
-            self._list_available_subtitles(video_id, talk_info)
-            return
-
-        thumbnail = talk_info['thumb']
-        if not thumbnail.startswith('http'):
-            thumbnail = 'http://' + thumbnail
+

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

-            'title': talk_info['title'].strip(),
-            'uploader': talk_info['speaker'],
-            'thumbnail': thumbnail,
+            'title': title,
+            'uploader': player_talk.get('speaker') or talk_info.get('speaker'),
+            'thumbnail': player_talk.get('thumb') or talk_info.get('thumb'),

             'description': self._og_search_description(webpage),
             'description': self._og_search_description(webpage),

-            'subtitles': video_subtitles,
+            'subtitles': self._get_subtitles(video_id, talk_info),

             'formats': formats,
             'formats': formats,

-            'duration': talk_info.get('duration'),
+            'duration': float_or_none(talk_info.get('duration')),
+            'view_count': int_or_none(data.get('viewed_count')),
+            'comment_count': int_or_none(
+                try_get(data, lambda x: x['comments']['count'])),
+            'tags': try_get(talk_info, lambda x: x['tags'], list),

         }
 
         }
 

-    def _get_available_subtitles(self, video_id, talk_info):
-        languages = [lang['languageCode'] for lang in talk_info.get('languages', [])]
-        if languages:
-            sub_lang_list = {}
-            for l in languages:
-                url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
-                sub_lang_list[l] = url
-            return sub_lang_list
-        else:
-            self._downloader.report_warning('video doesn\'t have subtitles')
-            return {}
+    def _get_subtitles(self, video_id, talk_info):
+        sub_lang_list = {}
+        for language in try_get(
+                talk_info,
+                (lambda x: x['downloads']['languages'],
+                 lambda x: x['languages']), list):
+            lang_code = language.get('languageCode') or language.get('ianaCode')
+            if not lang_code:
+                continue
+            sub_lang_list[lang_code] = [
+                {
+                    'url': 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/%s' % (video_id, lang_code, ext),
+                    'ext': ext,
+                }
+                for ext in ['ted', 'srt']
+            ]
+        return sub_lang_list

 
     def _watch_info(self, url, name):
         webpage = self._download_webpage(url, name)
 
         config_json = self._html_search_regex(
 
     def _watch_info(self, url, name):
         webpage = self._download_webpage(url, name)
 
         config_json = self._html_search_regex(

-            r"data-config='([^']+)", webpage, 'config')
-        config = json.loads(config_json)
+            r'"pages\.jwplayer"\s*,\s*({.+?})\s*\)\s*</script>',
+            webpage, 'config', default=None)
+        if not config_json:
+            embed_url = self._search_regex(
+                r"<iframe[^>]+class='pages-video-embed__video__object'[^>]+src='([^']+)'", webpage, 'embed url')
+            return self.url_result(self._proto_relative_url(embed_url))
+        config = json.loads(config_json)['config']

         video_url = config['video']['url']
         thumbnail = config.get('image', {}).get('url')
 
         video_url = config['video']['url']
         thumbnail = config.get('image', {}).get('url')