[ie/crunchyroll] Fix stream extraction (#10005)

[yt-dlp.git] / yt_dlp / extractor / mediasite.py

diff --git a/yt_dlp/extractor/mediasite.py b/yt_dlp/extractor/mediasite.py
index 30464bad0e745880ac2176fdc38704eb20308c70..7ea78ab69184ae2f6246f34f8def4dee7e782953 100644 (file)
--- a/yt_dlp/extractor/mediasite.py
+++ b/yt_dlp/extractor/mediasite.py
@@ -13,7 +13,7 @@
     str_or_none,
     try_call,
     try_get,
-    unescapeHTML,
+    smuggle_url,
     unsmuggle_url,
     url_or_none,
     urljoin,
@@ -25,6 +25,7 @@
 
 class MediasiteIE(InfoExtractor):
     _VALID_URL = r'(?xi)https?://[^/]+/Mediasite/(?:Play|Showcase/[^/#?]+/Presentation)/(?P<id>%s)(?P<query>\?[^#]+|)' % _ID_RE
+    _EMBED_REGEX = [r'(?xi)<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:(?:https?:)?//[^/]+)?/Mediasite/Play/%s(?:\?.*?)?)\1' % _ID_RE]
     _TESTS = [
         {
             'url': 'https://hitsmediaweb.h-its.org/mediasite/Play/2db6c271681e4f199af3c60d1f82869b1d',
@@ -112,13 +113,10 @@ class MediasiteIE(InfoExtractor):
         5: 'video3',
     }
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return [
-            unescapeHTML(mobj.group('url'))
-            for mobj in re.finditer(
-                r'(?xi)<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:(?:https?:)?//[^/]+)?/Mediasite/Play/%s(?:\?.*?)?)\1' % _ID_RE,
-                webpage)]
+    @classmethod
+    def _extract_embed_urls(cls, url, webpage):
+        for embed_url in super()._extract_embed_urls(url, webpage):
+            yield smuggle_url(embed_url, {'UrlReferrer': url})
 
     def __extract_slides(self, *, stream_id, snum, Stream, duration, images):
         slide_base_url = Stream['SlideBaseUrl']
@@ -173,7 +171,7 @@ def _real_extract(self, url):
         query = mobj.group('query')
 
         webpage, urlh = self._download_webpage_handle(url, resource_id)  # XXX: add UrlReferrer?
-        redirect_url = urlh.geturl()
+        redirect_url = urlh.url
 
         # XXX: might have also extracted UrlReferrer and QueryString from the html
         service_path = compat_urlparse.urljoin(redirect_url, self._html_search_regex(
@@ -266,8 +264,6 @@ def _real_extract(self, url):
                 })
             formats.extend(stream_formats)
 
-        self._sort_formats(formats)
-
         # XXX: Presentation['Presenters']
         # XXX: Presentation['Transcript']