upstream-archive/VoiceStudio
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(dub): close shared workflow review gaps

Browse Source
This commit is contained in:
Palash Debnath
2026-08-27 20:43:21 +05:30
parent 422dbd1313
commit e1a76f9aca
11 changed files with 228 additions and 59 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/api/routers/dub_core.py
+20 -3
View File
@@ -507,13 +507,28 @@ _ingest_gen = dub_pipeline.ingest_pipeline
#: container so a mislabelled video can't slip past the video-skipping branch.
_AUDIO_EXTS = {".wav", ".mp3", ".m4a", ".aac", ".flac", ".ogg", ".opus", ".wma"}
# Source-language choices exposed by the first-party dub UI. Keeping this an
# allow-list rejects language names and private-use BCP-47 tags before they are
# persisted as ASR overrides. Values are normalized to lowercase below.
_DUB_SOURCE_LANG_CODES = frozenset({
"af", "sq", "am", "ar", "hy", "az", "eu", "be", "bn", "bs", "bg",
"my", "ca", "cmn-hans", "cmn-hant", "hr", "cs", "da", "nl", "en",
"et", "fi", "fr", "gl", "ka", "de", "el", "gu", "ht", "ha", "haw",
"he", "hi", "hu", "is", "id", "it", "ja", "jw", "kn", "kk", "km",
"ko", "ku", "ky", "lo", "la", "lv", "lt", "mk", "ms", "ml", "mt",
"mi", "mr", "mn", "ne", "no", "ps", "fa", "pl", "pt", "pa", "ro",
"ru", "sm", "gd", "sr", "sn", "sd", "si", "sk", "sl", "so", "es",
"su", "sw", "sv", "tg", "ta", "te", "th", "tr", "uk", "ur", "uz",
"vi", "cy", "xh", "yi", "yo", "zu",
})
def _source_lang_override(value: str | None) -> str | None:
"""Normalize a user-selected source language; auto/und means detect."""
code = (value or "").strip().lower()
if code in {"", "auto", "und"}:
return None
if len(code) > 35 or not all(ch.isalnum() or ch == "-" for ch in code):
if code not in _DUB_SOURCE_LANG_CODES:
raise HTTPException(status_code=400, detail="Invalid source language code")
return code
@@ -553,6 +568,7 @@ async def dub_upload(
detail=f"Audio-only dubbing needs an audio file ({', '.join(sorted(_AUDIO_EXTS))}); got '{ext or 'no extension'}'.",
)
source_lang_override = _source_lang_override(source_lang)
os.makedirs(job_dir, exist_ok=True)
video_path = os.path.join(job_dir, f"original{ext}")
@@ -568,7 +584,7 @@ async def dub_upload(
"kind": "file",
"path": video_path,
"input_type": input_type,
"source_lang": _source_lang_override(source_lang),
"source_lang": source_lang_override,
},
filename,
)
@@ -592,6 +608,7 @@ async def dub_ingest_url(req: DubIngestUrlRequest, request: Request):
status_code=400,
detail="URL must start with http:// or https://. Paste a full video link (e.g. https://youtube.com/watch?v=…) or drop a local file instead.",
)
source_lang_override = _source_lang_override(req.source_lang)
try:
import yt_dlp # noqa: F401
@@ -627,7 +644,7 @@ async def dub_ingest_url(req: DubIngestUrlRequest, request: Request):
"fetch_subs": bool(req.fetch_subs),
"sub_langs": req.sub_langs or None,
"cookie_file": cookie_path,
"source_lang": _source_lang_override(req.source_lang),
"source_lang": source_lang_override,
}
try:
await task_manager.add_task(
backend/services/dub_pipeline.py
+19 -9
View File
@@ -755,35 +755,45 @@ async def _ensure_browser_playable_mp4_for_job(job_id: str, video_path: str) ->
run_proc = run_proc_factory(job_id)
ffmpeg_bin = find_ffmpeg()
async def attempt(cmd: list[str]) -> int:
try:
proc, _stdout, _stderr = await run_proc(cmd, timeout=1800.0)
return proc.returncode
except asyncio.CancelledError:
raise
except Exception as exc:
logger.warning(
"Browser-media normalization process failed for %s: %s",
log_safe(video_path),
log_safe(exc),
)
return 1
rc = 1
if not is_mp4:
proc, _stdout, _stderr = await run_proc(
rc = await attempt(
[
ffmpeg_bin, "-y", "-i", video_path,
"-c:v", "copy", "-c:a", "copy",
"-movflags", "+faststart", target,
],
timeout=1800.0,
]
)
rc = proc.returncode
if rc != 0 or not os.path.exists(target):
rc = 1
if rc != 0:
proc, _stdout, _stderr = await run_proc(
rc = await attempt(
[
ffmpeg_bin, "-y", "-i", video_path,
"-c:v", "libx264", "-preset", "veryfast", "-crf", "23",
"-pix_fmt", "yuv420p", "-c:a", "aac", "-b:a", "192k",
"-movflags", "+faststart", target,
],
timeout=1800.0,
]
)
rc = proc.returncode
if rc == 0 and os.path.exists(target) and target != video_path:
try:
os.remove(video_path)
except OSError:
pass
pass # Best effort: the normalized target is already complete.
return target
logger.warning(
"Could not transcode %s to browser-playable mp4 — the in-app "
backend/services/segmentation.py
+11 -7
View File
@@ -117,17 +117,21 @@ def _merge_segment_extra(target: Segment, incoming: Segment, *, prepend: bool) -
else joined(target_original, incoming_original)
)
target_translations = target.extra.get("translations")
incoming_translations = incoming.extra.get("translations")
if isinstance(target_translations, dict) or isinstance(incoming_translations, dict):
raw_target_translations = target.extra.get("translations")
raw_incoming_translations = incoming.extra.get("translations")
target_translations = raw_target_translations if isinstance(raw_target_translations, dict) else {}
incoming_translations = (
raw_incoming_translations if isinstance(raw_incoming_translations, dict) else {}
)
if target_translations or incoming_translations:
merged = {}
languages = {
*(target_translations or {}).keys(),
*(incoming_translations or {}).keys(),
*target_translations.keys(),
*incoming_translations.keys(),
}
for language in languages:
target_text = (target_translations or {}).get(language)
incoming_text = (incoming_translations or {}).get(language)
target_text = target_translations.get(language)
incoming_text = incoming_translations.get(language)
merged[language] = (
joined(incoming_text, target_text)
if prepend
backend/tests/test_dub_preview_upload.py
+4 -9
View File
@@ -12,11 +12,12 @@ from fastapi import UploadFile
async def test_preview_ffmpeg_does_not_block_event_loop(monkeypatch, tmp_path):
from api.routers import dub_core
started = threading.Event()
loop = asyncio.get_running_loop()
started = asyncio.Event()
release = threading.Event()
def slow_ffmpeg(*_args, **_kwargs):
started.set()
loop.call_soon_threadsafe(started.set)
assert release.wait(timeout=2)
monkeypatch.setattr(dub_core, "PREVIEW_DIR", str(tmp_path))
@@ -24,16 +25,10 @@ async def test_preview_ffmpeg_does_not_block_event_loop(monkeypatch, tmp_path):
monkeypatch.setattr(dub_core.subprocess, "run", slow_ffmpeg)
upload = UploadFile(filename="preview.mp4", file=io.BytesIO(b"video"))
loop = asyncio.get_running_loop()
before = loop.time()
task = asyncio.create_task(dub_core.preview_upload(upload))
try:
deadline = loop.time() + 1.0
while not started.is_set():
if loop.time() >= deadline:
pytest.fail("ffmpeg extraction did not start")
await asyncio.sleep(0.01)
await asyncio.sleep(0.05)
await asyncio.wait_for(started.wait(), timeout=1)
assert loop.time() - before < 0.5
finally:
release.set()
frontend/src/hooks/useDubWorkflow.js
+9 -7
View File
@@ -70,7 +70,6 @@ export default function useDubWorkflow({
const dubLang = useAppStore((s) => s.dubLang);
const dubLangCode = useAppStore((s) => s.dubLangCode);
const dubSourceLangCode = useAppStore((s) => s.dubSourceLangCode);
const setDubSourceLangCode = useAppStore((s) => s.setDubSourceLangCode);
const dubInstruct = useAppStore((s) => s.dubInstruct);
const setDubFilename = useAppStore((s) => s.setDubFilename);
const setDubDuration = useAppStore((s) => s.setDubDuration);
@@ -286,9 +285,6 @@ export default function useDubWorkflow({
const castSources = m.cast_sources || m.speaker_clones || {};
setDubSegments(applySpeakerCloneDefaults(normalized, castSources));
setDubTranscript(m.full_transcript || '');
if (useAppStore.getState().dubSourceLangCode === 'auto' && m.source_lang) {
setDubSourceLangCode(m.source_lang);
}
if (castSources && typeof castSources === 'object') {
setSpeakerClones(castSources);
}
@@ -365,7 +361,7 @@ export default function useDubWorkflow({
).then(reject, reject);
});
}),
[setDubSegments, setDubTranscript, setSpeakerClones, setDubSourceLangCode],
[setDubSegments, setDubTranscript, setSpeakerClones],
);
// ── SSE: wait for prep pipeline ──
@@ -860,12 +856,18 @@ export default function useDubWorkflow({
// that language rather than rendering a wrong-language track.
const handleTranslateAll = useCallback(
async (langOverride) => {
const options =
langOverride && typeof langOverride === 'object' && !('preventDefault' in langOverride)
? langOverride
: {};
const targetLang =
typeof langOverride === 'string' && langOverride ? langOverride : dubLangCode;
typeof langOverride === 'string' && langOverride
? langOverride
: options.langOverride || dubLangCode;
// Snapshot segments at call time: inside the multi-language loop the
// click-time closure is stale after the previous pick's translate pass.
const allSegments = useAppStore.getState().dubSegments;
const retryFailed = !!langOverride?.retryFailed;
const retryFailed = !!options.retryFailed;
const segs = retryFailed
? allSegments.filter((segment) => segment.translate_error)
: allSegments;
frontend/src/pages/DubTab.jsx
+24 -21
View File
@@ -322,28 +322,31 @@ export default function DubTab(props) {
// while still restoring the primary target in the editor afterwards. Each
// translation lands in segments[].translations[code], so Generate can
// reuse the complete maps without retranslating or losing another language.
const onTranslateClick = useCallback(async () => {
if (!multiLangMode) return handleTranslateAll();
if (multiBatchRunningRef.current) return false;
multiBatchRunningRef.current = true;
setMultiBatchBusy(true);
const { lang: primaryLanguage, code: primaryCode } = primaryTargetRef.current;
let allOk = true;
try {
for (const target of batchTargets) {
setDubLang(target.lang);
switchDubLangCode(target.code);
const ok = await handleTranslateAll(target.code);
if (!ok) allOk = false;
const onTranslateClick = useCallback(
async (options = {}) => {
if (!multiLangMode) return handleTranslateAll(options);
if (multiBatchRunningRef.current) return false;
multiBatchRunningRef.current = true;
setMultiBatchBusy(true);
const { lang: primaryLanguage, code: primaryCode } = primaryTargetRef.current;
let allOk = true;
try {
for (const target of batchTargets) {
setDubLang(target.lang);
switchDubLangCode(target.code);
const ok = await handleTranslateAll({ ...options, langOverride: target.code });
if (!ok) allOk = false;
}
} finally {
setDubLang(primaryLanguage);
switchDubLangCode(primaryCode);
multiBatchRunningRef.current = false;
setMultiBatchBusy(false);
}
} finally {
setDubLang(primaryLanguage);
switchDubLangCode(primaryCode);
multiBatchRunningRef.current = false;
setMultiBatchBusy(false);
}
return allOk;
}, [multiLangMode, batchTargets, handleTranslateAll, setDubLang, switchDubLangCode]);
return allOk;
},
[multiLangMode, batchTargets, handleTranslateAll, setDubLang, switchDubLangCode],
);
// Live ETA while generating — elapsed ticks each second; remaining is
// extrapolated from the current/total rate so it's only meaningful once
frontend/src/test/dubAsrInstallRecovery.test.jsx
+34
View File
@@ -117,6 +117,40 @@ describe('Dubbing missing-ASR recovery', () => {
expect(useAppStore.getState().dubJobId).not.toBe('job-kept-for-retry');
});
it('keeps auto detection selected across two consecutive jobs', async () => {
let uploadNumber = 0;
dubApi.dubUpload.mockImplementation(async () => {
uploadNumber += 1;
return { job_id: `job-${uploadNumber}`, task_id: `prep-${uploadNumber}` };
});
const { result } = renderWorkflow();
const runUpload = async (detectedLanguage) => {
const streamStart = streams.length;
let upload;
act(() => {
upload = result.current.handleDubUpload(
new File(['video'], `job-${uploadNumber + 1}.mp4`, { type: 'video/mp4' }),
);
});
await waitFor(() => expect(streams).toHaveLength(streamStart + 1));
streams[streamStart].emit('message', { type: 'ready' });
await waitFor(() => expect(streams).toHaveLength(streamStart + 2));
streams[streamStart + 1].emit('final', {
segments: [{ id: '1', text: 'hello' }],
source_lang: detectedLanguage,
});
streams[streamStart + 1].emit('done');
await act(async () => upload);
};
await runUpload('es');
await runUpload('de');
expect(dubApi.dubUpload.mock.calls.map((call) => call[2].sourceLang)).toEqual(['auto', 'auto']);
expect(useAppStore.getState().dubSourceLangCode).toBe('auto');
});
it('keeps the job, installs inline, then automatically retranscribes it', async () => {
const { result } = renderWorkflow();
let firstAttempt;
frontend/src/test/dubMultiLangGenerate.test.jsx
+23 -3
View File
@@ -14,14 +14,19 @@ import { useAppStore } from '../store';
// and on a failed translate: skip that pick's generate, keep going, report
// the skipped languages in a final toast.
const captured = vi.hoisted(() => ({ header: [] }));
const captured = vi.hoisted(() => ({ header: [], left: [] }));
vi.mock('../components/dub/DubHeader', () => ({
default: (props) => {
captured.header.push(props);
return null;
},
}));
vi.mock('../components/dub/DubLeftColumn', () => ({ default: () => null }));
vi.mock('../components/dub/DubLeftColumn', () => ({
default: (props) => {
captured.left.push(props);
return null;
},
}));
vi.mock('../components/dub/DubRightColumn', () => ({ default: () => null }));
vi.mock('../components/dub/DubFooter', () => ({ default: () => null }));
vi.mock('../components/dub/DubPipelineStepper', () => ({ default: () => null }));
@@ -101,6 +106,7 @@ const PICKS = [
{ lang: 'French', code: 'fr' },
{ lang: 'German', code: 'de' },
];
const EXPECTED_CODES = ['bn', ...PICKS.map((pick) => pick.code)];
/** Render DubTab in multi-lang mode and return { onGenerateClick, calls, mocks }. */
function setup({
@@ -110,7 +116,8 @@ function setup({
segments,
} = {}) {
const calls = [];
const handleTranslateAll = vi.fn(async (code) => {
const handleTranslateAll = vi.fn(async (arg) => {
const code = typeof arg === 'string' ? arg : arg?.langOverride;
calls.push(`translate:${code}`);
const ok = translateOk(code);
if (ok) {
@@ -155,6 +162,19 @@ describe('DubTab — multi-language generate translates each language first (P1.
beforeEach(() => {
useAppStore.setState(baseState, true);
captured.header.length = 0;
captured.left.length = 0;
});
it('forwards retry-only options through every language in the wrapper', async () => {
const { handleTranslateAll } = setup();
await act(async () => {
await captured.left.at(-1).handleTranslateAll({ retryFailed: true });
});
expect(handleTranslateAll.mock.calls.map(([options]) => options)).toEqual(
EXPECTED_CODES.map((langOverride) => ({ retryFailed: true, langOverride })),
);
});
afterEach(() => {
vi.restoreAllMocks();
tests/test_dub_browser_media.py
+24
View File
@@ -5,6 +5,8 @@ from __future__ import annotations
import asyncio
from types import SimpleNamespace
import pytest
from services import dub_pipeline as dp
@@ -123,3 +125,25 @@ def test_upload_normalization_propagates_cancellation_to_registered_process(tmp_
asyncio.run(cancel_normalization())
assert seen == {"job_id": "cancel-job", "timeout": 1800.0}
@pytest.mark.parametrize("failure", [RuntimeError("spawn failed"), asyncio.TimeoutError()])
def test_upload_normalization_failure_keeps_the_original_media(tmp_path, monkeypatch, failure):
source = tmp_path / "source.mp4"
source.write_bytes(b"source")
monkeypatch.setattr(dp, "_probe_codecs", lambda _path: ("vp9", "opus"))
monkeypatch.setattr(dp, "find_ffmpeg", lambda: "ffmpeg")
def factory(_job_id):
async def run_proc(_cmd, *, timeout):
assert timeout == 1800.0
raise failure
return run_proc
monkeypatch.setattr(dp, "run_proc_factory", factory)
result = asyncio.run(dp._ensure_browser_playable_mp4_for_job("failed-job", str(source)))
assert result == str(source)
assert source.exists()
tests/test_dub_cleanup_preserves_translations.py
+25
View File
@@ -37,3 +37,28 @@ def test_cleanup_preserves_editor_metadata_and_combines_translations():
}
assert cleaned[0]["profile_id"] == "voice-a"
assert cleaned[0]["translate_error"] == "retry me"
def test_cleanup_ignores_legacy_non_mapping_translations():
segments = [
{
"id": "a",
"start": 0.0,
"end": 2.0,
"text": "Hello.",
"speaker_id": "Speaker 1",
"translations": "legacy-corrupt-value",
},
{
"id": "b",
"start": 2.1,
"end": 2.4,
"text": "Again.",
"speaker_id": "Speaker 1",
"translations": {"es": "Otra vez."},
},
]
cleaned = clean_up_segments(segments)
assert cleaned[0]["translations"] == {"es": "Otra vez."}
tests/test_dub_export_unique.py
+35
View File
@@ -292,3 +292,38 @@ class TestAudioOnlyDubbing:
)
assert r.status_code == 400
assert "audio file" in r.json()["detail"].lower()
def test_ingest_requests_reject_unregistered_source_languages(self, app_client):
client, _dc, _dx, _tmp = app_client
form = client.post(
"/dub/upload",
files={"video": ("clip.mp4", b"video", "video/mp4")},
data={"source_lang": "english"},
)
json_response = client.post(
"/dub/ingest-url",
json={"url": "https://example.com/video", "source_lang": "x-123"},
)
assert form.status_code == 400
assert json_response.status_code == 400
assert form.json()["detail"] == "Invalid source language code"
assert json_response.json()["detail"] == "Invalid source language code"
def test_upload_accepts_a_registered_source_language(self, app_client, monkeypatch):
client, dc, _dx, _tmp = app_client
queued = []
async def add_task(*args):
queued.append(args)
monkeypatch.setattr(dc.task_manager, "add_task", add_task)
response = client.post(
"/dub/upload",
files={"video": ("clip.mp4", b"video", "video/mp4")},
data={"source_lang": "FR"},
)
assert response.status_code == 202
assert queued[0][5]["source_lang"] == "fr"