Updated version 2017.10.15.1 from 'upstream/2017.10.15.1'

[youtubedl] / youtube_dl / extractor / eagleplatform.py

diff --git a/youtube_dl/extractor/eagleplatform.py b/youtube_dl/extractor/eagleplatform.py
index 7bbf617d468f9d5295ee2e1123452679a21a7e4b..42789278e3a04753bcbce06ad62316c458708dc1 100644 (file)
--- a/youtube_dl/extractor/eagleplatform.py
+++ b/youtube_dl/extractor/eagleplatform.py
@@ -4,9 +4,14 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
 import re
 
 from .common import InfoExtractor

+from ..compat import (
+    compat_HTTPError,
+    compat_str,
+)

 from ..utils import (
     ExtractorError,
     int_or_none,
 from ..utils import (
     ExtractorError,
     int_or_none,

+    unsmuggle_url,

 )
 
 
 )
 
 


@@ -21,13 +26,13 @@ class EaglePlatformIE(InfoExtractor):
     _TESTS = [{
         # http://lenta.ru/news/2015/03/06/navalny/
         'url': 'http://lentaru.media.eagleplatform.com/index/player?player=new&record_id=227304&player_template_id=5201',
     _TESTS = [{
         # http://lenta.ru/news/2015/03/06/navalny/
         'url': 'http://lentaru.media.eagleplatform.com/index/player?player=new&record_id=227304&player_template_id=5201',

-        'md5': '70f5187fb620f2c1d503b3b22fd4efe3',
+        # Not checking MD5 as sometimes the direct HTTP link results in 404 and HLS is used

         'info_dict': {
             'id': '227304',
             'ext': 'mp4',
             'title': 'Навальный вышел на свободу',
             'description': 'md5:d97861ac9ae77377f3f20eaf9d04b4f5',
         'info_dict': {
             'id': '227304',
             'ext': 'mp4',
             'title': 'Навальный вышел на свободу',
             'description': 'md5:d97861ac9ae77377f3f20eaf9d04b4f5',

-            'thumbnail': 're:^https?://.*\.jpg$',
+            'thumbnail': r're:^https?://.*\.jpg$',

             'duration': 87,
             'view_count': int,
             'age_limit': 0,
             'duration': 87,
             'view_count': int,
             'age_limit': 0,


@@ -36,38 +41,106 @@ class EaglePlatformIE(InfoExtractor):
         # http://muz-tv.ru/play/7129/
         # http://media.clipyou.ru/index/player?record_id=12820&width=730&height=415&autoplay=true
         'url': 'eagleplatform:media.clipyou.ru:12820',
         # http://muz-tv.ru/play/7129/
         # http://media.clipyou.ru/index/player?record_id=12820&width=730&height=415&autoplay=true
         'url': 'eagleplatform:media.clipyou.ru:12820',

-        'md5': '90b26344ba442c8e44aa4cf8f301164a',
+        'md5': '358597369cf8ba56675c1df15e7af624',

         'info_dict': {
             'id': '12820',
             'ext': 'mp4',
             'title': "'O Sole Mio",
         'info_dict': {
             'id': '12820',
             'ext': 'mp4',
             'title': "'O Sole Mio",

-            'thumbnail': 're:^https?://.*\.jpg$',
+            'thumbnail': r're:^https?://.*\.jpg$',

             'duration': 216,
             'view_count': int,
         },
         'skip': 'Georestricted',
             'duration': 216,
             'view_count': int,
         },
         'skip': 'Georestricted',

+    }, {
+        # referrer protected video (https://tvrain.ru/lite/teleshow/kak_vse_nachinalos/namin-418921/)
+        'url': 'eagleplatform:tvrainru.media.eagleplatform.com:582306',
+        'only_matching': True,

     }]
 
     }]
 

+    @staticmethod
+    def _extract_url(webpage):
+        # Regular iframe embedding
+        mobj = re.search(
+            r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//.+?\.media\.eagleplatform\.com/index/player\?.+?)\1',
+            webpage)
+        if mobj is not None:
+            return mobj.group('url')
+        PLAYER_JS_RE = r'''
+                        <script[^>]+
+                            src=(?P<qjs>["\'])(?:https?:)?//(?P<host>(?:(?!(?P=qjs)).)+\.media\.eagleplatform\.com)/player/player\.js(?P=qjs)
+                        .+?
+                    '''
+        # "Basic usage" embedding (see http://dultonmedia.github.io/eplayer/)
+        mobj = re.search(
+            r'''(?xs)
+                    %s
+                    <div[^>]+
+                        class=(?P<qclass>["\'])eagleplayer(?P=qclass)[^>]+
+                        data-id=["\'](?P<id>\d+)
+            ''' % PLAYER_JS_RE, webpage)
+        if mobj is not None:
+            return 'eagleplatform:%(host)s:%(id)s' % mobj.groupdict()
+        # Generalization of "Javascript code usage", "Combined usage" and
+        # "Usage without attaching to DOM" embeddings (see
+        # http://dultonmedia.github.io/eplayer/)
+        mobj = re.search(
+            r'''(?xs)
+                    %s
+                    <script>
+                    .+?
+                    new\s+EaglePlayer\(
+                        (?:[^,]+\s*,\s*)?
+                        {
+                            .+?
+                            \bid\s*:\s*["\']?(?P<id>\d+)
+                            .+?
+                        }
+                    \s*\)
+                    .+?
+                    </script>
+            ''' % PLAYER_JS_RE, webpage)
+        if mobj is not None:
+            return 'eagleplatform:%(host)s:%(id)s' % mobj.groupdict()
+

     @staticmethod
     def _handle_error(response):
         status = int_or_none(response.get('status', 200))
         if status != 200:
             raise ExtractorError(' '.join(response['errors']), expected=True)
 
     @staticmethod
     def _handle_error(response):
         status = int_or_none(response.get('status', 200))
         if status != 200:
             raise ExtractorError(' '.join(response['errors']), expected=True)
 

-    def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata'):
-        response = super(EaglePlatformIE, self)._download_json(url_or_request, video_id, note)
-        self._handle_error(response)
+    def _download_json(self, url_or_request, video_id, *args, **kwargs):
+        try:
+            response = super(EaglePlatformIE, self)._download_json(
+                url_or_request, video_id, *args, **kwargs)
+        except ExtractorError as ee:
+            if isinstance(ee.cause, compat_HTTPError):
+                response = self._parse_json(ee.cause.read().decode('utf-8'), video_id)
+                self._handle_error(response)
+            raise

         return response
 
     def _get_video_url(self, url_or_request, video_id, note='Downloading JSON metadata'):
         return self._download_json(url_or_request, video_id, note)['data'][0]
 
     def _real_extract(self, url):
         return response
 
     def _get_video_url(self, url_or_request, video_id, note='Downloading JSON metadata'):
         return self._download_json(url_or_request, video_id, note)['data'][0]
 
     def _real_extract(self, url):

+        url, smuggled_data = unsmuggle_url(url, {})
+

         mobj = re.match(self._VALID_URL, url)
         host, video_id = mobj.group('custom_host') or mobj.group('host'), mobj.group('id')
 
         mobj = re.match(self._VALID_URL, url)
         host, video_id = mobj.group('custom_host') or mobj.group('host'), mobj.group('id')
 

+        headers = {}
+        query = {
+            'id': video_id,
+        }
+
+        referrer = smuggled_data.get('referrer')
+        if referrer:
+            headers['Referer'] = referrer
+            query['referrer'] = referrer
+

         player_data = self._download_json(
         player_data = self._download_json(

-            'http://%s/api/player_data?id=%s' % (host, video_id), video_id)
+            'http://%s/api/player_data' % host, video_id,
+            headers=headers, query=query)

 
         media = player_data['data']['playlist']['viewports'][0]['medialist'][0]
 
 
         media = player_data['data']['playlist']['viewports'][0]['medialist'][0]
 


@@ -84,17 +157,42 @@ class EaglePlatformIE(InfoExtractor):
 
         secure_m3u8 = self._proto_relative_url(media['sources']['secure_m3u8']['auto'], 'http:')
 
 
         secure_m3u8 = self._proto_relative_url(media['sources']['secure_m3u8']['auto'], 'http:')
 

+        formats = []
+

         m3u8_url = self._get_video_url(secure_m3u8, video_id, 'Downloading m3u8 JSON')
         m3u8_url = self._get_video_url(secure_m3u8, video_id, 'Downloading m3u8 JSON')

-        formats = self._extract_m3u8_formats(
-            m3u8_url, video_id,
-            'mp4', entry_protocol='m3u8_native', m3u8_id='hls')
+        m3u8_formats = self._extract_m3u8_formats(
+            m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native',
+            m3u8_id='hls', fatal=False)
+        formats.extend(m3u8_formats)
+
+        m3u8_formats_dict = {}
+        for f in m3u8_formats:
+            if f.get('height') is not None:
+                m3u8_formats_dict[f['height']] = f

 
 

-        mp4_url = self._get_video_url(
+        mp4_data = self._download_json(

             # Secure mp4 URL is constructed according to Player.prototype.mp4 from
             # http://lentaru.media.eagleplatform.com/player/player.js
             # Secure mp4 URL is constructed according to Player.prototype.mp4 from
             # http://lentaru.media.eagleplatform.com/player/player.js

-            re.sub(r'm3u8|hlsvod|hls|f4m', 'mp4', secure_m3u8),
-            video_id, 'Downloading mp4 JSON')
-        formats.append({'url': mp4_url, 'format_id': 'mp4'})
+            re.sub(r'm3u8|hlsvod|hls|f4m', 'mp4s', secure_m3u8),
+            video_id, 'Downloading mp4 JSON', fatal=False)
+        if mp4_data:
+            for format_id, format_url in mp4_data.get('data', {}).items():
+                if not isinstance(format_url, compat_str):
+                    continue
+                height = int_or_none(format_id)
+                if height is not None and m3u8_formats_dict.get(height):
+                    f = m3u8_formats_dict[height].copy()
+                    f.update({
+                        'format_id': f['format_id'].replace('hls', 'http'),
+                        'protocol': 'http',
+                    })
+                else:
+                    f = {
+                        'format_id': 'http-%s' % format_id,
+                        'height': int_or_none(format_id),
+                    }
+                f['url'] = format_url
+                formats.append(f)

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