[blip.tv] Add support for subtitles (#2274)

test/test_subtitles.py

@@ -10,6 +10,7 @@
 
 
 from youtube_dl.extractor import (
+    BlipTVIE,
     YoutubeIE,
     DailymotionIE,
     TEDIE,
@@ -202,5 +203,25 @@ def test_multiple_langs(self):
         for lang in langs:
             self.assertTrue(subtitles.get(lang) is not None, u'Subtitles for \'%s\' not extracted' % lang)
 
+
+class TestBlipTVSubtitles(BaseTestSubtitles):
+    url = 'http://blip.tv/a/a-6603250'
+    IE = BlipTVIE
+
+    def test_list_subtitles(self):
+        self.DL.expect_warning(u'Automatic Captions not supported by this server')
+        self.DL.params['listsubtitles'] = True
+        info_dict = self.getInfoDict()
+        self.assertEqual(info_dict, None)
+
+    def test_allsubtitles(self):
+        self.DL.expect_warning(u'Automatic Captions not supported by this server')
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertEqual(set(subtitles.keys()), set(['en']))
+        self.assertEqual(md5(subtitles['en']), '5b75c300af65fe4476dff79478bb93e4')
+
+
 if __name__ == '__main__':
     unittest.main()

youtube_dl/extractor/bliptv.py

@@ -6,6 +6,7 @@
 import socket
 
 from .common import InfoExtractor
+from .subtitles import SubtitlesInfoExtractor
 from ..utils import (
     compat_http_client,
     compat_str,
@@ -17,112 +18,124 @@
 )
 
 
-class BlipTVIE(InfoExtractor):
+class BlipTVIE(SubtitlesInfoExtractor):
     """Information extractor for blip.tv"""
 
-    _VALID_URL = r'^(?:https?://)?(?:\w+\.)?blip\.tv/((.+/)|(play/)|(api\.swf#))(.+)$'
+    _VALID_URL = r'https?://(?:\w+\.)?blip\.tv/((.+/)|(play/)|(api\.swf#))(?P<presumptive_id>.+)$'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://blip.tv/cbr/cbr-exclusive-gotham-city-imposters-bats-vs-jokerz-short-3-5796352',
-        'file': '5779306.mov',
         'md5': 'c6934ad0b6acf2bd920720ec888eb812',
         'info_dict': {
+            'id': '5779306',
+            'ext': 'mov',
             'upload_date': '20111205',
             'description': 'md5:9bc31f227219cde65e47eeec8d2dc596',
             'uploader': 'Comic Book Resources - CBR TV',
             'title': 'CBR EXCLUSIVE: "Gotham City Imposters" Bats VS Jokerz Short 3',
         }
-    }
-
-    def report_direct_download(self, title):
-        """Report information extraction."""
-        self.to_screen('%s: Direct download detected' % title)
+    }, {
+        # https://github.com/rg3/youtube-dl/pull/2274
+        'note': 'Video with subtitles',
+        'url': 'http://blip.tv/play/h6Uag5OEVgI.html',
+        'md5': '309f9d25b820b086ca163ffac8031806',
+        'info_dict': {
+            'id': '6586561',
+            'ext': 'mp4',
+            'uploader': 'Red vs. Blue',
+            'description': 'One-Zero-One',
+            'upload_date': '20130614',
+            'title': 'Red vs. Blue Season 11 Episode 1',
+        }
+    }]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        if mobj is None:
-            raise ExtractorError('Invalid URL: %s' % url)
+        presumptive_id = mobj.group('presumptive_id')
 
         # See https://github.com/rg3/youtube-dl/issues/857
-        embed_mobj = re.search(r'^(?:https?://)?(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)([a-zA-Z0-9]+)', url)
+        embed_mobj = re.match(r'https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)([a-zA-Z0-9]+)', url)
         if embed_mobj:
             info_url = 'http://blip.tv/play/%s.x?p=1' % embed_mobj.group(1)
             info_page = self._download_webpage(info_url, embed_mobj.group(1))
-            video_id = self._search_regex(r'data-episode-id="(\d+)', info_page,  'video_id')
+            video_id = self._search_regex(
+                r'data-episode-id="([0-9]+)', info_page, 'video_id')
             return self.url_result('http://blip.tv/a/a-' + video_id, 'BlipTV')
         
-        if '?' in url:
-            cchar = '&'
-        else:
-            cchar = '?'
+        cchar = '&' if '?' in url else '?'
         json_url = url + cchar + 'skin=json&version=2&no_wrap=1'
         request = compat_urllib_request.Request(json_url)
         request.add_header('User-Agent', 'iTunes/10.6.1')
-        self.report_extraction(mobj.group(1))
-        urlh = self._request_webpage(request, None, False,
-            'unable to download video info webpage')
+        json_data = self._download_json(request, video_id=presumptive_id)
 
-        try:
-            json_code_bytes = urlh.read()
-            json_code = json_code_bytes.decode('utf-8')
-        except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
-            raise ExtractorError('Unable to read video info webpage: %s' % compat_str(err))
+        if 'Post' in json_data:
+            data = json_data['Post']
+        else:
+            data = json_data
 
-        try:
-            json_data = json.loads(json_code)
-            if 'Post' in json_data:
-                data = json_data['Post']
-            else:
-                data = json_data
-
-            upload_date = datetime.datetime.strptime(data['datestamp'], '%m-%d-%y %H:%M%p').strftime('%Y%m%d')
-            formats = []
-            if 'additionalMedia' in data:
-                for f in sorted(data['additionalMedia'], key=lambda f: int(f['media_height'])):
-                    if not int(f['media_width']): # filter m3u8
-                        continue
-                    formats.append({
-                        'url': f['url'],
-                        'format_id': f['role'],
-                        'width': int(f['media_width']),
-                        'height': int(f['media_height']),
-                    })
-            else:
+        video_id = compat_str(data['item_id'])
+        upload_date = datetime.datetime.strptime(data['datestamp'], '%m-%d-%y %H:%M%p').strftime('%Y%m%d')
+        subtitles = {}
+        formats = []
+        if 'additionalMedia' in data:
+            for f in data['additionalMedia']:
+                if f.get('file_type_srt') == 1:
+                    LANGS = {
+                        'english': 'en',
+                    }
+                    lang = f['role'].rpartition('-')[-1].strip().lower()
+                    langcode = LANGS.get(lang, lang)
+                    subtitles[langcode] = f['url']
+                    continue
+                if not int(f['media_width']):  # filter m3u8
+                    continue
                 formats.append({
-                    'url': data['media']['url'],
-                    'width': int(data['media']['width']),
-                    'height': int(data['media']['height']),
+                    'url': f['url'],
+                    'format_id': f['role'],
+                    'width': int(f['media_width']),
+                    'height': int(f['media_height']),
                 })
+        else:
+            formats.append({
+                'url': data['media']['url'],
+                'width': int(data['media']['width']),
+                'height': int(data['media']['height']),
+            })
+        self._sort_formats(formats)
 
-            self._sort_formats(formats)
+        # subtitles
+        video_subtitles = self.extract_subtitles(video_id, subtitles)
+        if self._downloader.params.get('listsubtitles', False):
+            self._list_available_subtitles(video_id, subtitles)
+            return
 
-            return {
-                'id': compat_str(data['item_id']),
-                'uploader': data['display_name'],
-                'upload_date': upload_date,
-                'title': data['title'],
-                'thumbnail': data['thumbnailUrl'],
-                'description': data['description'],
-                'user_agent': 'iTunes/10.6.1',
-                'formats': formats,
-            }
-        except (ValueError, KeyError) as err:
-            raise ExtractorError('Unable to parse video information: %s' % repr(err))
+        return {
+            'id': video_id,
+            'uploader': data['display_name'],
+            'upload_date': upload_date,
+            'title': data['title'],
+            'thumbnail': data['thumbnailUrl'],
+            'description': data['description'],
+            'user_agent': 'iTunes/10.6.1',
+            'formats': formats,
+            'subtitles': video_subtitles,
+        }
+
+    def _download_subtitle_url(self, sub_lang, url):
+        # For some weird reason, blip.tv serves a video instead of subtitles
+        # when we request with a common UA
+        req = compat_urllib_request.Request(url)
+        req.add_header('Youtubedl-user-agent', 'youtube-dl')
+        return self._download_webpage(req, None, note=False)
 
 
 class BlipTVUserIE(InfoExtractor):
-    """Information Extractor for blip.tv users."""
-
     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?blip\.tv/)|bliptvuser:)([^/]+)/*$'
     _PAGE_SIZE = 12
     IE_NAME = 'blip.tv:user'
 
     def _real_extract(self, url):
-        # Extract username
         mobj = re.match(self._VALID_URL, url)
-        if mobj is None:
-            raise ExtractorError('Invalid URL: %s' % url)
-
         username = mobj.group(1)
 
         page_base = 'http://m.blip.tv/pr/show_get_full_episode_list?users_id=%s&lite=0&esi=1'
@@ -131,7 +144,6 @@ def _real_extract(self, url):
         mobj = re.search(r'data-users-id="([^"]+)"', page)
         page_base = page_base % mobj.group(1)
 
-
         # Download video ids using BlipTV Ajax calls. Result size per
         # query is limited (currently to 12 videos) so we need to query
         # page by page until there are no video ids - it means we got
@@ -142,8 +154,8 @@ def _real_extract(self, url):
 
         while True:
             url = page_base + "&page=" + str(pagenum)
-            page = self._download_webpage(url, username,
-                                          'Downloading video ids from page %d' % pagenum)
+            page = self._download_webpage(
+                url, username, 'Downloading video ids from page %d' % pagenum)
 
             # Extract video identifiers
             ids_in_page = []
@@ -167,4 +179,4 @@ def _real_extract(self, url):
 
         urls = ['http://blip.tv/%s' % video_id for video_id in video_ids]
         url_entries = [self.url_result(vurl, 'BlipTV') for vurl in urls]
-        return [self.playlist_result(url_entries, playlist_title = username)]
+        return [self.playlist_result(url_entries, playlist_title=username)]

youtube_dl/extractor/subtitles.py

@@ -62,24 +62,30 @@ def extract_subtitles(self, video_id, webpage):
                 subtitles[sub_lang] = subtitle
         return subtitles
 
+    def _download_subtitle_url(self, sub_lang, url):
+        return self._download_webpage(url, None, note=False)
+
     def _request_subtitle_url(self, sub_lang, url):
         """ makes the http request for the subtitle """
         try:
-            sub = self._download_webpage(url, None, note=False)
+            return self._download_subtitle_url(sub_lang, url)
         except ExtractorError as err:
             self._downloader.report_warning(u'unable to download video subtitles for %s: %s' % (sub_lang, compat_str(err)))
             return
         if not sub:
             self._downloader.report_warning(u'Did not fetch video subtitles')
             return
-        return sub
 
     def _get_available_subtitles(self, video_id, webpage):
         """
         returns {sub_lang: url} or {} if not available
         Must be redefined by the subclasses
         """
-        pass
+
+        # By default, allow implementations to simply pass in the result
+        assert isinstance(webpage, dict), \
+            '_get_available_subtitles not implemented'
+        return webpage
 
     def _get_available_automatic_caption(self, video_id, webpage):
         """