Skip to content

The open-source AI voice studio. Clone, dictate, create.

test(export): cover MP3 export endpoint and make the m4b round-trip probe version-proof

jamiepine1927d0607b72
diff --git a/backend/tests/test_generation_export_mp3.py b/backend/tests/test_generation_export_mp3.py
new file mode 100644
index 00000000..b5448aa8
--- /dev/null
+++ b/backend/tests/test_generation_export_mp3.py
@@ -0,0 +1,122 @@
+"""Tests for ``GET /history/{id}/export-audio?format=`` and the shared ffmpeg helper.
+
+The real transcode runs only when ffmpeg/ffprobe are installed; the missing-ffmpeg
+path is exercised everywhere by patching the lookup.
+"""
+
+import json
+import shutil
+import subprocess
+import sys
+from pathlib import Path
+
+import numpy as np
+import pytest
+from fastapi import FastAPI
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+from starlette.testclient import TestClient
+
+sys.path.insert(0, str(Path(__file__).parent.parent.parent))
+
+from backend import config
+from backend.database import Base, Generation, VoiceProfile, get_db
+from backend.routes.history import router as history_router
+from backend.utils import ffmpeg as ffmpeg_util
+from backend.utils.audio import save_audio
+
+HAS_FFMPEG = shutil.which("ffmpeg") is not None and shutil.which("ffprobe") is not None
+
+
+def client(tmp_path, monkeypatch):
+ monkeypatch.setattr(config, "_data_dir", tmp_path)
+ (tmp_path / "generations").mkdir()
+ tone = 0.2 * np.sin(np.linspace(0, 2 * np.pi * 440 * 2, 24000 * 2)).astype(np.float32)
+ save_audio(tone, str(tmp_path / "generations" / "gen-1.wav"), 24000)
+
+ engine = create_engine(
+ f"sqlite:///{tmp_path / 'test.db'}", connect_args={"check_same_thread": False}
+ )
+ Base.metadata.create_all(bind=engine)
+ session_local = sessionmaker(autocommit=False, autoflush=False, bind=engine)
+ session = session_local()
+ session.add(VoiceProfile(id="profile-1", name="Test Profile"))
+ session.add(
+ Generation(
+ id="gen-1",
+ profile_id="profile-1",
+ text="Hello there, this is a test.",
+ audio_path="generations/gen-1.wav",
+ status="completed",
+ )
+ )
+ session.commit()
+ session.close()
+
+ app = FastAPI()
+ app.include_router(history_router)
+
+ def override_get_db():
+ db = session_local()
+ try:
+ yield db
+ finally:
+ db.close()
+
+ app.dependency_overrides[get_db] = override_get_db
+ return TestClient(app)
+
+
+def test_default_export_is_unchanged_wav(client, tmp_path):
+ response = client.get("/history/gen-1/export-audio")
+ assert response.status_code == 200
+ assert response.headers["content-type"].startswith("audio/wav")
+ assert 'filename="Hello there this is a test-gen-1.wav"' in response.headers["content-disposition"]
+ assert response.content == (tmp_path / "generations" / "gen-1.wav").read_bytes()
+
+
+def test_unknown_format_is_400(client):
+ assert client.get("/history/gen-1/export-audio?format=ogg").status_code == 400
+
+
+def test_missing_ffmpeg_is_503_with_install_hint(client, monkeypatch):
+ monkeypatch.setattr(ffmpeg_util, "find_ffmpeg", lambda name="ffmpeg": None)
+ response = client.get("/history/gen-1/export-audio?format=mp3")
+ assert response.status_code == 503
+ assert response.json()["detail"] == ffmpeg_util.MISSING_MESSAGE
+
+
+def test_encode_audio_rejects_unknown_format(tmp_path):
+ with pytest.raises(ValueError, match="Unsupported export format"):
+ ffmpeg_util.encode_audio(tmp_path / "in.wav", tmp_path / "out.ogg", "ogg")
+
+
+def test_find_ffmpeg_honours_override_dir(tmp_path, monkeypatch):
+ fake = tmp_path / "ffmpeg"
+ fake.write_text("#!/bin/sh\n")
+ monkeypatch.setenv("VOICEBOX_FFMPEG_DIR", str(tmp_path))
+ assert ffmpeg_util.find_ffmpeg() == str(fake)
+
+
[email protected](not HAS_FFMPEG, reason="ffmpeg/ffprobe not installed")
+def test_mp3_export_is_real_mp3(client, tmp_path):
+ response = client.get("/history/gen-1/export-audio?format=mp3")
+ assert response.status_code == 200
+ assert response.headers["content-type"] == "audio/mpeg"
+ assert 'filename="Hello there this is a test-gen-1.mp3"' in response.headers["content-disposition"]
+
+ out = tmp_path / "probe.mp3"
+ out.write_bytes(response.content)
+ probe = subprocess.run(
+ ["ffprobe", "-v", "error", "-show_streams", "-show_format", "-of", "json", str(out)],
+ capture_output=True,
+ check=True,
+ text=True,
+ )
+ info = json.loads(probe.stdout)
+ assert info["format"]["format_name"] == "mp3"
+ assert info["streams"][0]["codec_name"] == "mp3"
+ assert abs(float(info["format"]["duration"]) - 2.0) < 0.2
+ # Nothing was cached beside the WAV.
+ assert sorted(p.name for p in (tmp_path / "generations").iterdir()) == ["gen-1.wav"]
diff --git a/backend/tests/test_story_export.py b/backend/tests/test_story_export.py
index 3351adaa..41be8b8c 100644
--- a/backend/tests/test_story_export.py
+++ b/backend/tests/test_story_export.py
@@ -7,6 +7,7 @@ under a shutil.which gate so the suite stays green on CI runners without ffmpeg.
from __future__ import annotations
+import json
import shutil
import subprocess
from pathlib import Path
@@ -153,15 +154,13 @@ class TestFFmpegEncodeIntegration:
"error",
"-show_chapters",
"-of",
- "default=noprint_wrappers=1",
+ "json",
str(out_path),
],
capture_output=True,
check=True,
text=True,
)
- chunks = probe.stdout.split("[CHAPTER]")
- # ffprobe prints one [CHAPTER] block per chapter; expect two.
- assert sum(1 for c in chunks if c.strip()) == 2
- assert "Opening" in probe.stdout
- assert "Closing" in probe.stdout
+ probed = json.loads(probe.stdout)["chapters"]
+ assert [c["tags"]["title"] for c in probed] == ["Opening", "Closing"]
+ assert [(c["start"], c["end"]) for c in probed] == [(0, 4000), (4000, 12000)]