⚡️(back) redirect video url instead of streaming on /transcript-source

The transcript-source video api endpoint was streaming the video
content. Instead of doing this we can return a permanent redirect
response to the video url. Peertube runner is able to manage this
redirection, it releases pressure on our servers and will directly use a
CDN URL.

Author: lunika

CHANGELOG.md

@@ -12,6 +12,7 @@ Versioning](https://semver.org/spec/v2.0.0.html).
 
 - Allow to configure transcoding resolutions
 - Upgrade to python 3.12
+- Redirect video url instead of streaming it on /transcript-source
 
 ## [5.3.1] - 2024-11-07
 

src/backend/marsha/core/api/video.py

@@ -11,16 +11,15 @@
 from django.core.exceptions import ValidationError
 from django.db import IntegrityError, OperationalError, transaction
 from django.db.models import F, Func, Q, Value
-from django.http import Http404, StreamingHttpResponse
-from django.shortcuts import get_object_or_404
+from django.http import Http404
+from django.shortcuts import get_object_or_404, redirect
 from django.urls import reverse
 from django.utils import timezone
 from django.utils.module_loading import import_string
 
 from boto3.exceptions import Boto3Error
 import django_filters
 from django_peertube_runner_connector.models import RunnerJob
-import requests
 from rest_framework import filters, status, viewsets
 from rest_framework.decorators import action
 from rest_framework.exceptions import APIException, MethodNotAllowed
@@ -1283,24 +1282,6 @@ def transcript_source(self, request, pk=None):
                 {"detail": "No video source available for this video."},
                 status=HTTPStatus.NOT_FOUND,
             )
-        video_url = video_urls.get(min(video_urls.keys()))
+        video_url = video_urls.get(max(video_urls.keys()))
 
-        # stream the video source
-        try:
-            response = requests.get(
-                video_url,
-                stream=True,
-                timeout=settings.TRANSCRIPTION_VIDEO_SOURCE_TIMEOUT,
-            )
-            response.raise_for_status()
-        except (requests.HTTPError, requests.RequestException) as err:
-            return Response(
-                {"detail": f"Error occurred: {err}"}, status=HTTPStatus.BAD_REQUEST
-            )
-
-        return StreamingHttpResponse(
-            response.iter_content(
-                chunk_size=settings.TRANSCRIPTION_VIDEO_SOURCE_CHUNK_SIZE
-            ),
-            content_type=response.headers["Content-Type"],
-        )
+        return redirect(video_url, permanent=True)

src/backend/marsha/core/tests/api/video/test_transcript_source.py

@@ -4,8 +4,6 @@
 from http import HTTPStatus
 
 from django_peertube_runner_connector.factories import RunnerJobFactory
-import requests
-import responses
 from rest_framework.test import APITestCase
 
 from marsha.core import factories
@@ -26,30 +24,21 @@ def setUp(self):
         )
         self.url = f"/api/videos/{self.video.pk}/transcript-source/"
         serializer = VideoSerializer(self.video)
-        self.video_url = serializer.data.get("urls").get("mp4").get(144)
+        self.video_url = serializer.data.get("urls").get("mp4").get(720)
         runner_job = RunnerJobFactory()
         self.runner_job_data = {
             "runnerToken": runner_job.runner.runnerToken,
             "jobToken": runner_job.processingJobToken,
         }
 
-    @responses.activate
     def test_api_video_transcript_source(self):
-        """The API should return the video source."""
-        responses.get(
-            url=self.video_url,
-            body=b"video content",
-            status=200,
-            content_type="video/mp4",
-        )
+        """The API should redirect to the video url."""
 
         response = self.client.post(self.url, data=self.runner_job_data)
 
-        self.assertEqual(response.status_code, HTTPStatus.OK)
-        self.assertEqual(response["Content-Type"], "video/mp4")
-        self.assertEqual(b"".join(response.streaming_content), b"video content")
+        self.assertEqual(response.status_code, HTTPStatus.MOVED_PERMANENTLY)
+        self.assertEqual(response["Location"], self.video_url)
 
-    @responses.activate
     def test_api_video_transcript_source_missing_video_url(self):
         """The API should return a 404 if the video source is not available."""
         self.video.resolutions = []
@@ -63,28 +52,8 @@ def test_api_video_transcript_source_missing_video_url(self):
             status_code=HTTPStatus.NOT_FOUND,
         )
 
-    @responses.activate
-    def test_api_video_transcript_source_request_failure(self):
-        """The API should return a 400 if the request to the video source fails."""
-        responses.get(
-            url=self.video_url,
-            body=requests.RequestException(),
-            status=500,
-        )
-
-        response = self.client.post(self.url, data=self.runner_job_data)
-
-        self.assertEqual(response.status_code, HTTPStatus.BAD_REQUEST)
-
-    @responses.activate
     def test_api_video_transcript_source_no_runner(self):
         """The API should return a 403 if the request is not from a runner job."""
-        responses.get(
-            url=self.video_url,
-            body=b"video content",
-            status=200,
-            content_type="video/mp4",
-        )
         data = {
             "runnerToken": "unknown-runner-token",
             "jobToken": "unknown-job-token",