[youtube] Improve SAPISID cookie handling (closes #393) (#395)

Author: colethedj

diff --git a/yt_dlp/extractor/youtube.py b/yt_dlp/extractor/youtube.py
index e2a174ae12e4a5c175d4c7b1bc99bf1be7139028..f88c2f727dcb85b114e8fb0e5471f9d1381bfaeb 100644 (file)
--- a/yt_dlp/extractor/youtube.py
+++ b/yt_dlp/extractor/youtube.py
@@ -301,12 +301,22 @@ def _real_initialize(self):
     _YT_INITIAL_BOUNDARY_RE = r'(?:var\s+meta|</script|\n)'
 
     def _generate_sapisidhash_header(self):
-        sapisid_cookie = self._get_cookies('https://www.youtube.com').get('SAPISID')
+        # Sometimes SAPISID cookie isn't present but __Secure-3PAPISID is.
+        # See: https://github.com/yt-dlp/yt-dlp/issues/393
+        yt_cookies = self._get_cookies('https://www.youtube.com')
+        sapisid_cookie = dict_get(
+            yt_cookies, ('__Secure-3PAPISID', 'SAPISID'))
         if sapisid_cookie is None:
             return
         time_now = round(time.time())
-        sapisidhash = hashlib.sha1((str(time_now) + " " + sapisid_cookie.value + " " + "https://www.youtube.com").encode("utf-8")).hexdigest()
-        return "SAPISIDHASH %s_%s" % (time_now, sapisidhash)
+        # SAPISID cookie is required if not already present
+        if not yt_cookies.get('SAPISID'):
+            self._set_cookie(
+                '.youtube.com', 'SAPISID', sapisid_cookie.value, secure=True, expire_time=time_now + 3600)
+        # SAPISIDHASH algorithm from https://stackoverflow.com/a/32065323
+        sapisidhash = hashlib.sha1(
+            f'{time_now} {sapisid_cookie.value} https://www.youtube.com'.encode('utf-8')).hexdigest()
+        return f'SAPISIDHASH {time_now}_{sapisidhash}'
 
     def _call_api(self, ep, query, video_id, fatal=True, headers=None,
                   note='Downloading API JSON', errnote='Unable to download API page',