Implement Phase 6: MeTube download integration

- MeTubeClient encapsulates all MeTube HTTP/Socket.IO calls (verified against
  the real MeTube source: /add returns no job id, GET /history gives a queue
  snapshot for reconciliation, filenames arrive already relative, percent is
  a 0-100 float)
- download_jobs table + service: request/dedup active downloads, apply live
  Socket.IO events (added/updated/completed/canceled/cleared) matched by
  canonical YouTube URL, safe relative-path -> public media URL construction
- Reconciliation on startup against MeTube's live queue/done state (section 19):
  non-terminal jobs recovered where possible, else marked "unknown"; already
  completed jobs are left untouched
- POST/GET /api/videos/{id}/download(-status), recheck-local; feed/video
  detail now report real local availability instead of a stub
- Frontend: download button with live status polling (queued/downloading %/
  postprocessing/completed/failed+retry), local <video> playback with
  YouTube fallback on playback error
- health.py now delegates to MeTubeClient (single place for MeTube calls)

26 new backend tests (63 total). Verified live: Socket.IO connects
successfully to the real MeTube instance on deploy.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
This commit is contained in:
vrubelroman 2026-09-16 18:56:17 +00:00
parent 0ed20bb838
commit fe16c08daa
27 changed files with 1125 additions and 48 deletions

View file

@ -17,8 +17,9 @@ def client(monkeypatch, db_session):
monkeypatch.setattr(
google_oauth, "build_authorization_url", lambda: ("https://accounts.google.com/fake", "fixed-state")
)
with TestClient(app) as test_client:
yield test_client
# Deliberately not using `with TestClient(app)`: that runs the app's
# lifespan, which would try to reach the real MeTube instance and DB.
yield TestClient(app)
del app.dependency_overrides[get_db]

View file

@ -14,8 +14,9 @@ def client(db_session):
app.dependency_overrides[get_db] = _get_db_override
app.dependency_overrides[require_session] = lambda: None
with TestClient(app) as test_client:
yield test_client
# Deliberately not using `with TestClient(app)`: that runs the app's
# lifespan, which would try to reach the real MeTube instance and DB.
yield TestClient(app)
del app.dependency_overrides[get_db]
del app.dependency_overrides[require_session]

243
tests/test_download_jobs.py Normal file
View file

@ -0,0 +1,243 @@
import json
import pytest
from app.models.channel import Channel
from app.models.download_job import DownloadJob
from app.models.video import Video
from app.services import download_jobs
def _seed_video(db_session, youtube_video_id="vid1", youtube_channel_id="chanA"):
channel = Channel(youtube_channel_id=youtube_channel_id, title="Channel", subscribed=True)
db_session.add(channel)
db_session.commit()
from datetime import datetime, timezone
video = Video(
youtube_video_id=youtube_video_id,
channel_id=channel.id,
title="Video",
published_at=datetime(2026, 9, 10, tzinfo=timezone.utc),
youtube_url=f"https://www.youtube.com/watch?v={youtube_video_id}",
)
db_session.add(video)
db_session.commit()
db_session.refresh(video)
return video
def test_request_download_enqueues_and_creates_job(monkeypatch, db_session):
video = _seed_video(db_session)
calls = {}
def fake_enqueue(self, youtube_url, custom_name_prefix):
calls["youtube_url"] = youtube_url
calls["custom_name_prefix"] = custom_name_prefix
return {"status": "ok"}
monkeypatch.setattr("app.services.metube_client.MeTubeClient.enqueue_video", fake_enqueue)
job = download_jobs.request_download(db_session, video)
assert job.status == "queued"
assert calls["youtube_url"] == video.youtube_url
assert calls["custom_name_prefix"] == video.youtube_video_id
def test_request_download_is_idempotent_while_active(monkeypatch, db_session):
video = _seed_video(db_session)
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.enqueue_video", lambda self, u, p: {"status": "ok"}
)
job1 = download_jobs.request_download(db_session, video)
job2 = download_jobs.request_download(db_session, video)
assert job1.id == job2.id
assert db_session.query(DownloadJob).count() == 1
def test_request_download_raises_on_metube_error(monkeypatch, db_session):
video = _seed_video(db_session)
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.enqueue_video",
lambda self, u, p: {"status": "error", "msg": "boom"},
)
with pytest.raises(download_jobs.MeTubeRejected):
download_jobs.request_download(db_session, video)
assert db_session.query(DownloadJob).count() == 0
def test_request_download_allows_retry_after_failure(monkeypatch, db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="failed", error_message="oops")
db_session.add(job)
db_session.commit()
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.enqueue_video", lambda self, u, p: {"status": "ok"}
)
new_job = download_jobs.request_download(db_session, video)
assert new_job.id != job.id
assert db_session.query(DownloadJob).count() == 2
@pytest.mark.asyncio
async def test_handle_metube_event_added_and_updated(db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="queued")
db_session.add(job)
db_session.commit()
added_payload = json.dumps({"id": "vid1.vid1", "url": video.youtube_url, "status": "pending"})
await download_jobs.handle_metube_event(db_session, "added", added_payload)
db_session.refresh(job)
assert job.metube_job_id == "vid1.vid1"
assert job.status == "queued"
updating_payload = json.dumps(
{"id": "vid1.vid1", "url": video.youtube_url, "status": "downloading", "percent": 42.5}
)
await download_jobs.handle_metube_event(db_session, "updated", updating_payload)
db_session.refresh(job)
assert job.status == "downloading"
assert job.progress_percent == 42
assert job.started_at is not None
@pytest.mark.asyncio
async def test_handle_metube_event_completed_builds_media_url(monkeypatch, db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="downloading", metube_job_id="vid1.vid1")
db_session.add(job)
db_session.commit()
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.build_media_url",
lambda self, filename: f"http://metube.local/download/{filename}",
)
payload = json.dumps(
{"id": "vid1.vid1", "url": video.youtube_url, "status": "finished", "filename": "vid1.vid1.mp4"}
)
await download_jobs.handle_metube_event(db_session, "completed", payload)
db_session.refresh(job)
assert job.status == "completed"
assert job.progress_percent == 100
assert job.metube_filename == "vid1.vid1.mp4"
assert job.media_url == "http://metube.local/download/vid1.vid1.mp4"
assert job.completed_at is not None
@pytest.mark.asyncio
async def test_handle_metube_event_error_marks_failed(db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="downloading", metube_job_id="vid1.vid1")
db_session.add(job)
db_session.commit()
payload = json.dumps({"id": "vid1.vid1", "url": video.youtube_url, "status": "error", "msg": "network blip"})
await download_jobs.handle_metube_event(db_session, "updated", payload)
db_session.refresh(job)
assert job.status == "failed"
assert job.error_message == "network blip"
@pytest.mark.asyncio
async def test_handle_metube_event_ignores_unrelated_download(db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="queued")
db_session.add(job)
db_session.commit()
payload = json.dumps(
{"id": "other.other", "url": "https://www.youtube.com/watch?v=other", "status": "finished"}
)
await download_jobs.handle_metube_event(db_session, "completed", payload)
db_session.refresh(job)
assert job.status == "queued"
@pytest.mark.asyncio
async def test_handle_metube_event_canceled_marks_failed(db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="downloading", metube_job_id="vid1.vid1")
db_session.add(job)
db_session.commit()
await download_jobs.handle_metube_event(db_session, "canceled", json.dumps("vid1.vid1"))
db_session.refresh(job)
assert job.status == "failed"
def test_reconcile_marks_unknown_when_history_unavailable(monkeypatch, db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="downloading")
db_session.add(job)
db_session.commit()
def raise_error(self):
raise RuntimeError("connection refused")
monkeypatch.setattr("app.services.metube_client.MeTubeClient.fetch_history", raise_error)
download_jobs.reconcile_on_startup(db_session)
db_session.refresh(job)
assert job.status == "unknown"
def test_reconcile_restores_completed_from_history(monkeypatch, db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="downloading")
db_session.add(job)
db_session.commit()
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.fetch_history",
lambda self: {
"queue": [],
"pending": [],
"done": [{"id": "vid1.vid1", "url": video.youtube_url, "status": "finished", "filename": "f.mp4"}],
},
)
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.build_media_url",
lambda self, filename: f"http://metube.local/download/{filename}",
)
download_jobs.reconcile_on_startup(db_session)
db_session.refresh(job)
assert job.status == "completed"
assert job.media_url == "http://metube.local/download/f.mp4"
def test_reconcile_leaves_completed_jobs_untouched(monkeypatch, db_session):
video = _seed_video(db_session)
job = DownloadJob(video_id=video.id, status="completed", media_url="http://x/f.mp4")
db_session.add(job)
db_session.commit()
def fail_if_called(self):
raise AssertionError("should not fetch history for completed jobs")
monkeypatch.setattr("app.services.metube_client.MeTubeClient.fetch_history", fail_if_called)
download_jobs.reconcile_on_startup(db_session)
db_session.refresh(job)
assert job.status == "completed"
assert job.media_url == "http://x/f.mp4"

View file

@ -19,8 +19,9 @@ def client(db_session):
app.dependency_overrides[get_db] = _get_db_override
app.dependency_overrides[require_session] = lambda: None
with TestClient(app) as test_client:
yield test_client
# Deliberately not using `with TestClient(app)`: that runs the app's
# lifespan, which would try to reach the real MeTube instance and DB.
yield TestClient(app)
del app.dependency_overrides[get_db]
del app.dependency_overrides[require_session]

View file

@ -21,8 +21,7 @@ def override_get_db():
def test_health_ok():
with patch("app.api.health.httpx.get") as mock_get:
mock_get.return_value = MagicMock(status_code=200)
with patch("app.services.metube_client.MeTubeClient.health", return_value=True):
client = TestClient(app)
response = client.get("/api/health")
@ -34,7 +33,7 @@ def test_health_ok():
def test_health_metube_unreachable():
with patch("app.api.health.httpx.get", side_effect=ConnectionError):
with patch("app.services.metube_client.MeTubeClient.health", return_value=False):
client = TestClient(app)
response = client.get("/api/health")

View file

@ -0,0 +1,33 @@
import pytest
from app.services.metube_client import MeTubeClient
@pytest.fixture
def client():
return MeTubeClient()
def test_build_media_url_encodes_spaces(client):
url = client.build_media_url("some file.mp4")
assert url == f"{client.public_base_url}/download/some%20file.mp4"
def test_build_media_url_strips_absolute_download_dir_prefix(client):
filename = f"{client.download_dir}/nested/video.mp4"
url = client.build_media_url(filename)
assert url == f"{client.public_base_url}/download/nested/video.mp4"
def test_build_media_url_rejects_path_traversal(client):
assert client.build_media_url("../../etc/passwd") is None
def test_build_media_url_none_for_empty(client):
assert client.build_media_url("") is None
assert client.build_media_url(None) is None
def test_build_media_url_encodes_each_segment(client):
url = client.build_media_url("dir with space/file#1.mp4")
assert url == f"{client.public_base_url}/download/dir%20with%20space/file%231.mp4"

View file

@ -19,8 +19,10 @@ def client(db_session):
app.dependency_overrides[get_db] = _get_db_override
app.dependency_overrides[require_session] = lambda: None
with TestClient(app) as test_client:
yield test_client
# Deliberately not using `with TestClient(app)`: that runs the app's
# lifespan, which would try to reach the real MeTube instance and DB
# (see AGENTS.md: MeTube network calls in tests must be opt-in only).
yield TestClient(app)
del app.dependency_overrides[get_db]
del app.dependency_overrides[require_session]
@ -60,3 +62,99 @@ def test_get_video_by_youtube_id(client, db_session):
def test_get_video_not_found(client):
resp = client.get("/api/videos/missing")
assert resp.status_code == 404
def _seed_single_video(db_session, youtube_video_id="vid1"):
channel = Channel(youtube_channel_id="chanA", title="Channel A", subscribed=True)
db_session.add(channel)
db_session.commit()
video = Video(
youtube_video_id=youtube_video_id,
channel_id=channel.id,
title="Video One",
published_at=datetime(2026, 9, 10, tzinfo=timezone.utc),
youtube_url=f"https://www.youtube.com/watch?v={youtube_video_id}",
)
db_session.add(video)
db_session.commit()
db_session.refresh(video)
return video
def test_download_video_enqueues(client, db_session, monkeypatch):
video = _seed_single_video(db_session)
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.enqueue_video", lambda self, u, p: {"status": "ok"}
)
resp = client.post(f"/api/videos/{video.youtube_video_id}/download")
assert resp.status_code == 200
assert resp.json()["status"] == "queued"
def test_download_video_metube_rejects(client, db_session, monkeypatch):
video = _seed_single_video(db_session)
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.enqueue_video",
lambda self, u, p: {"status": "error", "msg": "bad url"},
)
resp = client.post(f"/api/videos/{video.youtube_video_id}/download")
assert resp.status_code == 502
def test_download_video_second_click_returns_same_job(client, db_session, monkeypatch):
video = _seed_single_video(db_session)
monkeypatch.setattr(
"app.services.metube_client.MeTubeClient.enqueue_video", lambda self, u, p: {"status": "ok"}
)
r1 = client.post(f"/api/videos/{video.youtube_video_id}/download").json()
r2 = client.post(f"/api/videos/{video.youtube_video_id}/download").json()
assert r1["requested_at"] == r2["requested_at"]
from app.models.download_job import DownloadJob
assert db_session.query(DownloadJob).count() == 1
def test_download_status_not_downloaded(client, db_session):
video = _seed_single_video(db_session)
resp = client.get(f"/api/videos/{video.youtube_video_id}/download-status")
assert resp.json()["status"] == "not_downloaded"
def test_recheck_local_downgrades_to_unknown_when_media_missing(client, db_session, monkeypatch):
from app.models.download_job import DownloadJob
video = _seed_single_video(db_session)
job = DownloadJob(video_id=video.id, status="completed", media_url="http://metube.local/download/f.mp4")
db_session.add(job)
db_session.commit()
monkeypatch.setattr("app.services.metube_client.MeTubeClient.check_media", lambda self, url: False)
resp = client.post(f"/api/videos/{video.youtube_video_id}/recheck-local")
assert resp.status_code == 200
assert resp.json()["status"] == "unknown"
def test_recheck_local_keeps_completed_when_media_reachable(client, db_session, monkeypatch):
from app.models.download_job import DownloadJob
video = _seed_single_video(db_session)
job = DownloadJob(video_id=video.id, status="completed", media_url="http://metube.local/download/f.mp4")
db_session.add(job)
db_session.commit()
monkeypatch.setattr("app.services.metube_client.MeTubeClient.check_media", lambda self, url: True)
resp = client.post(f"/api/videos/{video.youtube_video_id}/recheck-local")
assert resp.status_code == 200
assert resp.json()["status"] == "completed"