Merge pull request #132 from jdepoix/feature/ISSUE-131

added error message if YouTube requests fail

youtube_transcript_api/__init__.py

@@ -12,4 +12,5 @@ from ._errors import (
     CookiePathInvalid,
     CookiesInvalid,
     FailedToCreateConsentCookie,
+    YouTubeRequestFailed,
 )

youtube_transcript_api/_errors.py

@@ -35,6 +35,20 @@ class CouldNotRetrieveTranscript(Exception):
         return self.CAUSE_MESSAGE
 
 
+class YouTubeRequestFailed(CouldNotRetrieveTranscript):
+    CAUSE_MESSAGE = 'Request to YouTube failed: {reason}'
+
+    def __init__(self, video_id, http_error):
+        self.reason = str(http_error)
+        super(YouTubeRequestFailed, self).__init__(video_id)
+
+    @property
+    def cause(self):
+        return self.CAUSE_MESSAGE.format(
+            reason=self.reason,
+        )
+
+
 class VideoUnavailable(CouldNotRetrieveTranscript):
     CAUSE_MESSAGE = 'The video is no longer available'
 

youtube_transcript_api/_transcripts.py

@@ -11,10 +11,13 @@ from xml.etree import ElementTree
 
 import re
 
+from requests import HTTPError
+
 from ._html_unescaping import unescape
 from ._errors import (
     VideoUnavailable,
     TooManyRequests,
+    YouTubeRequestFailed,
     NoTranscriptFound,
     TranscriptsDisabled,
     NotTranslatable,
@@ -25,6 +28,14 @@ from ._errors import (
 from ._settings import WATCH_URL
 
 
+def _raise_http_errors(response, video_id):
+    try:
+        response.raise_for_status()
+        return response
+    except HTTPError as error:
+        raise YouTubeRequestFailed(error, video_id)
+
+
 class TranscriptListFetcher(object):
     def __init__(self, http_client):
         self._http_client = http_client
@@ -72,7 +83,8 @@ class TranscriptListFetcher(object):
         return html
 
     def _fetch_html(self, video_id):
-        return self._http_client.get(WATCH_URL.format(video_id=video_id)).text.replace(
+        response = self._http_client.get(WATCH_URL.format(video_id=video_id))
+        return _raise_http_errors(response, video_id).text.replace(
             '\\u0026', '&'
         ).replace(
             '\\', ''
@@ -273,8 +285,9 @@ class Transcript(object):
         :return: a list of dictionaries containing the 'text', 'start' and 'duration' keys
         :rtype [{'text': str, 'start': float, 'end': float}]:
         """
+        response = self._http_client.get(self._url)
         return _TranscriptParser().parse(
-            self._http_client.get(self._url).text
+            _raise_http_errors(response, self.video_id).text,
         )
 
     def __str__(self):

youtube_transcript_api/test/test_api.py

@@ -19,6 +19,7 @@ from youtube_transcript_api import (
     CookiePathInvalid,
     CookiesInvalid,
     FailedToCreateConsentCookie,
+    YouTubeRequestFailed,
 )
 
 
@@ -174,6 +175,16 @@ class TestYouTubeTranscriptApi(TestCase):
         with self.assertRaises(VideoUnavailable):
             YouTubeTranscriptApi.get_transcript('abc')
 
+    def test_get_transcript__exception_if_youtube_request_fails(self):
+        httpretty.register_uri(
+            httpretty.GET,
+            'https://www.youtube.com/watch',
+            status=500
+        )
+
+        with self.assertRaises(YouTubeRequestFailed):
+            YouTubeTranscriptApi.get_transcript('abc')
+
     def test_get_transcript__exception_if_youtube_request_limit_reached(self):
         httpretty.register_uri(
             httpretty.GET,