Merge pull request #8436 from edx/waheed/tnl2122-fix-download-transcript-error-youtube-server

transcript name param url

cms/djangoapps/contentstore/tests/test_transcripts_utils.py

@@ -270,6 +270,107 @@ class TestDownloadYoutubeSubs(ModuleStoreTestCase):
 
         self.clear_sub_content(good_youtube_sub)
 
+    @patch('xmodule.video_module.transcripts_utils.requests.get')
+    def test_get_transcript_name_youtube_server_success(self, mock_get):
+        """
+        Get transcript name from transcript_list fetch from youtube server api
+        depends on language code, default language in YOUTUBE Text Api is "en"
+        """
+        youtube_text_api = copy.deepcopy(settings.YOUTUBE['TEXT_API'])
+        youtube_text_api['params']['v'] = 'dummy_video_id'
+        response_success = """
+        <transcript_list>
+            <track id="1" name="Custom" lang_code="en" />
+            <track id="0" name="Custom1" lang_code="en-GB"/>
+        </transcript_list>
+        """
+        mock_get.return_value = Mock(status_code=200, text=response_success, content=response_success)
+
+        transcript_name = transcripts_utils.youtube_video_transcript_name(youtube_text_api)
+        self.assertEqual(transcript_name, 'Custom')
+
+    @patch('xmodule.video_module.transcripts_utils.requests.get')
+    def test_get_transcript_name_youtube_server_no_transcripts(self, mock_get):
+        """
+        When there are no transcripts of video transcript name will be None
+        """
+        youtube_text_api = copy.deepcopy(settings.YOUTUBE['TEXT_API'])
+        youtube_text_api['params']['v'] = 'dummy_video_id'
+        response_success = "<transcript_list></transcript_list>"
+        mock_get.return_value = Mock(status_code=200, text=response_success, content=response_success)
+
+        transcript_name = transcripts_utils.youtube_video_transcript_name(youtube_text_api)
+        self.assertIsNone(transcript_name)
+
+    @patch('xmodule.video_module.transcripts_utils.requests.get')
+    def test_get_transcript_name_youtube_server_language_not_exist(self, mock_get):
+        """
+        When the language does not exist in transcript_list transcript name will be None
+        """
+        youtube_text_api = copy.deepcopy(settings.YOUTUBE['TEXT_API'])
+        youtube_text_api['params']['v'] = 'dummy_video_id'
+        youtube_text_api['params']['lang'] = 'abc'
+        response_success = """
+        <transcript_list>
+            <track id="1" name="Custom" lang_code="en" />
+            <track id="0" name="Custom1" lang_code="en-GB"/>
+        </transcript_list>
+        """
+        mock_get.return_value = Mock(status_code=200, text=response_success, content=response_success)
+
+        transcript_name = transcripts_utils.youtube_video_transcript_name(youtube_text_api)
+        self.assertIsNone(transcript_name)
+
+    def mocked_requests_get(*args, **kwargs):
+        """
+        This method will be used by the mock to replace requests.get
+        """
+        # pylint: disable=no-method-argument
+        response_transcript_list = """
+        <transcript_list>
+            <track id="1" name="Custom" lang_code="en" />
+            <track id="0" name="Custom1" lang_code="en-GB"/>
+        </transcript_list>
+        """
+        response_transcript = textwrap.dedent("""
+        <transcript>
+            <text start="0" dur="0.27"></text>
+            <text start="0.27" dur="2.45">Test text 1.</text>
+            <text start="2.72">Test text 2.</text>
+            <text start="5.43" dur="1.73">Test text 3.</text>
+        </transcript>
+        """)
+
+        if kwargs == {'params': {'lang': 'en', 'v': 'good_id_2'}}:
+            return Mock(status_code=200, text='')
+        elif kwargs == {'params': {'type': 'list', 'v': 'good_id_2'}}:
+            return Mock(status_code=200, text=response_transcript_list, content=response_transcript_list)
+        elif kwargs == {'params': {'lang': 'en', 'v': 'good_id_2', 'name': 'Custom'}}:
+            return Mock(status_code=200, text=response_transcript, content=response_transcript)
+
+        return Mock(status_code=404, text='')
+
+    @patch('xmodule.video_module.transcripts_utils.requests.get', side_effect=mocked_requests_get)
+    def test_downloading_subs_using_transcript_name(self, mock_get):
+        """
+        Download transcript using transcript name in url
+        """
+        good_youtube_sub = 'good_id_2'
+        self.clear_sub_content(good_youtube_sub)
+
+        transcripts_utils.download_youtube_subs(good_youtube_sub, self.course, settings)
+        mock_get.assert_any_call(
+            'http://video.google.com/timedtext',
+            params={'lang': 'en', 'v': 'good_id_2', 'name': 'Custom'}
+        )
+
+        # Check asset status after import of transcript.
+        filename = 'subs_{0}.srt.sjson'.format(good_youtube_sub)
+        content_location = StaticContent.compute_location(self.course.id, filename)
+        self.assertTrue(contentstore().find(content_location))
+
+        self.clear_sub_content(good_youtube_sub)
+
 
 class TestGenerateSubsFromSource(TestDownloadYoutubeSubs):
     """Tests for `generate_subs_from_source` function."""

common/lib/xmodule/xmodule/video_module/transcripts_utils.py

@@ -94,7 +94,32 @@ def save_subs_to_store(subs, subs_id, item, language='en'):
     return save_to_store(filedata, filename, 'application/json', item.location)
 
 
-def get_transcripts_from_youtube(youtube_id, settings, i18n):
+def youtube_video_transcript_name(youtube_text_api):
+    """
+    Get the transcript name from available transcripts of video
+    with respect to language from youtube server
+    """
+    # pylint: disable=no-member
+    utf8_parser = etree.XMLParser(encoding='utf-8')
+
+    transcripts_param = {'type': 'list', 'v': youtube_text_api['params']['v']}
+    lang = youtube_text_api['params']['lang']
+    # get list of transcripts of specific video
+    # url-form
+    # http://video.google.com/timedtext?type=list&v={VideoId}
+    youtube_response = requests.get('http://' + youtube_text_api['url'], params=transcripts_param)
+    if youtube_response.status_code == 200 and youtube_response.text:
+        # pylint: disable=no-member
+        youtube_data = etree.fromstring(youtube_response.content, parser=utf8_parser)
+        # iterate all transcripts information from youtube server
+        for element in youtube_data:
+            # search specific language code such as 'en' in transcripts info list
+            if element.tag == 'track' and element.get('lang_code', '') == lang:
+                return element.get('name')
+    return None
+
+
+def get_transcripts_from_youtube(youtube_id, settings, i18n, youtube_transcript_name=''):
     """
     Gets transcripts from youtube for youtube_id.
 
@@ -109,6 +134,12 @@ def get_transcripts_from_youtube(youtube_id, settings, i18n):
 
     youtube_text_api = copy.deepcopy(settings.YOUTUBE['TEXT_API'])
     youtube_text_api['params']['v'] = youtube_id
+    # if the transcript name is not empty on youtube server we have to pass
+    # name param in url in order to get transcript
+    # example http://video.google.com/timedtext?lang=en&v={VideoId}&name={transcript_name}
+    youtube_transcript_name = youtube_video_transcript_name(youtube_text_api)
+    if youtube_transcript_name:
+        youtube_text_api['params']['name'] = youtube_transcript_name
     data = requests.get('http://' + youtube_text_api['url'], params=youtube_text_api['params'])
 
     if data.status_code != 200 or not data.text: