Merge pull request #12909 from remitamine/raw-sub

[youtube-dl] / youtube_dl / extractor / youtube.py

diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index 5a6b735a09b87b839f01af91db40514170066604..ad2e933ee4e34c9ebdb982ca66278e6e4c4a06b0 100644 (file)
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -16,6 +16,7 @@ from ..jsinterp import JSInterpreter
 from ..swfinterp import SWFInterpreter
 from ..compat import (
     compat_chr,
+    compat_kwargs,
     compat_parse_qs,
     compat_urllib_parse_unquote,
     compat_urllib_parse_unquote_plus,
@@ -245,6 +246,11 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
 
         return True
 
+    def _download_webpage(self, *args, **kwargs):
+        kwargs.setdefault('query', {})['disable_polymer'] = 'true'
+        return super(YoutubeBaseInfoExtractor, self)._download_webpage(
+            *args, **compat_kwargs(kwargs))
+
     def _real_initialize(self):
         if self._downloader is None:
             return
@@ -1368,6 +1374,43 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             playback_url, video_id, 'Marking watched',
             'Unable to mark watched', fatal=False)
 
+    @staticmethod
+    def _extract_urls(webpage):
+        # Embedded YouTube player
+        entries = [
+            unescapeHTML(mobj.group('url'))
+            for mobj in re.finditer(r'''(?x)
+            (?:
+                <iframe[^>]+?src=|
+                data-video-url=|
+                <embed[^>]+?src=|
+                embedSWF\(?:\s*|
+                <object[^>]+data=|
+                new\s+SWFObject\(
+            )
+            (["\'])
+                (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
+                (?:embed|v|p)/.+?)
+            \1''', webpage)]
+
+        # lazyYT YouTube embed
+        entries.extend(list(map(
+            unescapeHTML,
+            re.findall(r'class="lazyYT" data-youtube-id="([^"]+)"', webpage))))
+
+        # Wordpress "YouTube Video Importer" plugin
+        matches = re.findall(r'''(?x)<div[^>]+
+            class=(?P<q1>[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+
+            data-video_id=(?P<q2>[\'"])([^\'"]+)(?P=q2)''', webpage)
+        entries.extend(m[-1] for m in matches)
+
+        return entries
+
+    @staticmethod
+    def _extract_url(webpage):
+        urls = YoutubeIE._extract_urls(webpage)
+        return urls[0] if urls else None
+
     @classmethod
     def extract_id(cls, url):
         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
@@ -1665,10 +1708,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         if not upload_date:
             upload_date = self._search_regex(
                 [r'(?s)id="eow-date.*?>(.*?)</span>',
-                 r'id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live|Started) on (.+?)</strong>'],
+                 r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']'],
                 video_webpage, 'upload date', default=None)
-            if upload_date:
-                upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
         upload_date = unified_strdate(upload_date)
 
         video_license = self._html_search_regex(
@@ -2054,7 +2095,7 @@ class YoutubePlaylistIE(YoutubePlaylistBaseInfoExtractor):
                      |
                         (%(playlist_id)s)
                      )""" % {'playlist_id': YoutubeBaseInfoExtractor._PLAYLIST_ID_RE}
-    _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&disable_polymer=true'
+    _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s'
     _VIDEO_RE = r'href="\s*/watch\?v=(?P<id>[0-9A-Za-z_-]{11})&amp;[^"]*?index=(?P<index>\d+)(?:[^>]+>(?P<title>[^<]+))?'
     IE_NAME = 'youtube:playlist'
     _TESTS = [{