debian/README.source: Add my preference with respect to patches.

[youtubedl] / youtube_dl / extractor / espn.py

diff --git a/youtube_dl/extractor/espn.py b/youtube_dl/extractor/espn.py
index e6f8f0337fa562335b577db3d1cee2d75684fe31..7a743606836517736e858411e626a5e2a502890e 100644 (file)
--- a/youtube_dl/extractor/espn.py
+++ b/youtube_dl/extractor/espn.py
@@ -1,24 +1,149 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    unified_timestamp,
+)
 
 
 class ESPNIE(InfoExtractor):
-    _VALID_URL = r'https?://espn\.go\.com/(?:[^/]+/)*(?P<id>[^/]+)'
-    _WORKING = False
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:
+                            (?:(?:\w+\.)+)?espn\.go|
+                            (?:www\.)?espn
+                        )\.com/
+                        (?:
+                            (?:
+                                video/clip|
+                                watch/player
+                            )
+                            (?:
+                                \?.*?\bid=|
+                                /_/id/
+                            )
+                        )
+                        (?P<id>\d+)
+                    '''
+
     _TESTS = [{
         'url': 'http://espn.go.com/video/clip?id=10365079',
         'info_dict': {
-            'id': 'FkYWtmazr6Ed8xmvILvKLWjd4QvYZpzG',
+            'id': '10365079',
+            'ext': 'mp4',
+            'title': '30 for 30 Shorts: Judging Jewell',
+            'description': 'md5:39370c2e016cb4ecf498ffe75bef7f0f',
+            'timestamp': 1390936111,
+            'upload_date': '20140128',
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://broadband.espn.go.com/video/clip?id=18910086',
+        'info_dict': {
+            'id': '18910086',
             'ext': 'mp4',
-            'title': 'dm_140128_30for30Shorts___JudgingJewellv2',
-            'description': '',
+            'title': 'Kyrie spins around defender for two',
+            'description': 'md5:2b0f5bae9616d26fba8808350f0d2b9b',
+            'timestamp': 1489539155,
+            'upload_date': '20170315',
         },
         'params': {
-            # m3u8 download
             'skip_download': True,
         },
+        'expected_warnings': ['Unable to download f4m manifest'],
+    }, {
+        'url': 'http://nonredline.sports.espn.go.com/video/clip?id=19744672',
+        'only_matching': True,
+    }, {
+        'url': 'https://cdn.espn.go.com/video/clip/_/id/19771774',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.espn.com/watch/player?id=19141491',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.espn.com/watch/player?bucketId=257&id=19505875',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.espn.com/watch/player/_/id/19141491',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.espn.com/video/clip?id=10365079',
+        'only_matching': True,
     }, {
+        'url': 'http://www.espn.com/video/clip/_/id/17989860',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        clip = self._download_json(
+            'http://api-app.espn.com/v1/video/clips/%s' % video_id,
+            video_id)['videos'][0]
+
+        title = clip['headline']
+
+        format_urls = set()
+        formats = []
+
+        def traverse_source(source, base_source_id=None):
+            for source_id, source in source.items():
+                if isinstance(source, compat_str):
+                    extract_source(source, base_source_id)
+                elif isinstance(source, dict):
+                    traverse_source(
+                        source,
+                        '%s-%s' % (base_source_id, source_id)
+                        if base_source_id else source_id)
+
+        def extract_source(source_url, source_id=None):
+            if source_url in format_urls:
+                return
+            format_urls.add(source_url)
+            ext = determine_ext(source_url)
+            if ext == 'smil':
+                formats.extend(self._extract_smil_formats(
+                    source_url, video_id, fatal=False))
+            elif ext == 'f4m':
+                formats.extend(self._extract_f4m_formats(
+                    source_url, video_id, f4m_id=source_id, fatal=False))
+            elif ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    source_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id=source_id, fatal=False))
+            else:
+                formats.append({
+                    'url': source_url,
+                    'format_id': source_id,
+                })
+
+        traverse_source(clip['links']['source'])
+        self._sort_formats(formats)
+
+        description = clip.get('caption') or clip.get('description')
+        thumbnail = clip.get('thumbnail')
+        duration = int_or_none(clip.get('duration'))
+        timestamp = unified_timestamp(clip.get('originalPublishDate'))
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnail,
+            'timestamp': timestamp,
+            'duration': duration,
+            'formats': formats,
+        }
+
+
+class ESPNArticleIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:espn\.go|(?:www\.)?espn)\.com/(?:[^/]+/)*(?P<id>[^/]+)'
+    _TESTS = [{
         'url': 'https://espn.go.com/video/iframe/twitter/?cms=espn&id=10365079',
         'only_matching': True,
     }, {
@@ -35,21 +160,18 @@ class ESPNIE(InfoExtractor):
         'only_matching': True,
     }]
 
+    @classmethod
+    def suitable(cls, url):
+        return False if ESPNIE.suitable(url) else super(ESPNArticleIE, cls).suitable(url)
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
         webpage = self._download_webpage(url, video_id)
 
         video_id = self._search_regex(
-            r'class="video-play-button"[^>]+data-id="(\d+)',
-            webpage, 'video id')
-
-        player = self._download_webpage(
-            'https://espn.go.com/video/iframe/twitter/?id=%s' % video_id, video_id)
-
-        pcode = self._search_regex(
-            r'["\']pcode=([^"\']+)["\']', player, 'pcode')
+            r'class=(["\']).*?video-play-button.*?\1[^>]+data-id=["\'](?P<id>\d+)',
+            webpage, 'video id', group='id')
 
         return self.url_result(
-            'ooyalaexternal:espn:%s:%s' % (video_id, pcode),
-            'OoyalaExternal')
+            'http://espn.go.com/video/clip?id=%s' % video_id, ESPNIE.ie_key())