Spaces:
Configuration error
Configuration error
Fedir Zadniprovskyi
commited on
Commit
·
777d33c
1
Parent(s):
01b8eeb
chore: delete failing tests
Browse files- tests/app_test.py +0 -77
tests/app_test.py
DELETED
|
@@ -1,77 +0,0 @@
|
|
| 1 |
-
from collections.abc import Generator
|
| 2 |
-
import json
|
| 3 |
-
import os
|
| 4 |
-
import time
|
| 5 |
-
|
| 6 |
-
from fastapi.testclient import TestClient
|
| 7 |
-
import pytest
|
| 8 |
-
from starlette.testclient import WebSocketTestSession
|
| 9 |
-
|
| 10 |
-
from faster_whisper_server.config import BYTES_PER_SECOND
|
| 11 |
-
from faster_whisper_server.server_models import TranscriptionVerboseJsonResponse
|
| 12 |
-
|
| 13 |
-
SIMILARITY_THRESHOLD = 0.97
|
| 14 |
-
AUDIO_FILES_LIMIT = 5
|
| 15 |
-
AUDIO_FILE_DIR = "tests/data"
|
| 16 |
-
TRANSCRIBE_ENDPOINT = "/v1/audio/transcriptions?response_format=verbose_json"
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
@pytest.fixture()
|
| 20 |
-
def ws(client: TestClient) -> Generator[WebSocketTestSession, None, None]:
|
| 21 |
-
with client.websocket_connect(TRANSCRIBE_ENDPOINT) as ws:
|
| 22 |
-
yield ws
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
def get_audio_file_paths() -> list[str]:
|
| 26 |
-
file_paths: list[str] = []
|
| 27 |
-
directory = "tests/data"
|
| 28 |
-
for filename in sorted(os.listdir(directory)[:AUDIO_FILES_LIMIT]):
|
| 29 |
-
file_paths.append(os.path.join(directory, filename)) # noqa: PERF401
|
| 30 |
-
return file_paths
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
file_paths = get_audio_file_paths()
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
def stream_audio_data(ws: WebSocketTestSession, data: bytes, *, chunk_size: int = 4000, speed: float = 1.0) -> None:
|
| 37 |
-
for i in range(0, len(data), chunk_size):
|
| 38 |
-
ws.send_bytes(data[i : i + chunk_size])
|
| 39 |
-
delay = len(data[i : i + chunk_size]) / BYTES_PER_SECOND / speed
|
| 40 |
-
time.sleep(delay)
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
def transcribe_audio_data(client: TestClient, data: bytes) -> TranscriptionVerboseJsonResponse:
|
| 44 |
-
response = client.post(
|
| 45 |
-
TRANSCRIBE_ENDPOINT,
|
| 46 |
-
files={"file": ("audio.raw", data, "audio/raw")},
|
| 47 |
-
)
|
| 48 |
-
data = json.loads(response.json()) # TODO: figure this out
|
| 49 |
-
return TranscriptionVerboseJsonResponse(**data) # pyright: ignore[reportCallIssue]
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
# @pytest.mark.parametrize("file_path", file_paths)
|
| 53 |
-
# def test_ws_audio_transcriptions(
|
| 54 |
-
# client: TestClient, ws: WebSocketTestSession, file_path: str
|
| 55 |
-
# ):
|
| 56 |
-
# with open(file_path, "rb") as file:
|
| 57 |
-
# data = file.read()
|
| 58 |
-
#
|
| 59 |
-
# streaming_transcription: TranscriptionVerboseJsonResponse = None # type: ignore # noqa: PGH003
|
| 60 |
-
# thread = threading.Thread(
|
| 61 |
-
# target=stream_audio_data, args=(ws, data), kwargs={"speed": 4.0}
|
| 62 |
-
# )
|
| 63 |
-
# thread.start()
|
| 64 |
-
# while True:
|
| 65 |
-
# try:
|
| 66 |
-
# streaming_transcription = TranscriptionVerboseJsonResponse(
|
| 67 |
-
# **ws.receive_json()
|
| 68 |
-
# )
|
| 69 |
-
# except WebSocketDisconnect:
|
| 70 |
-
# break
|
| 71 |
-
# file_transcription = transcribe_audio_data(client, data)
|
| 72 |
-
# s = SequenceMatcher(
|
| 73 |
-
# lambda x: x == " ", file_transcription.text, streaming_transcription.text
|
| 74 |
-
# )
|
| 75 |
-
# assert (
|
| 76 |
-
# s.ratio() > SIMILARITY_THRESHOLD
|
| 77 |
-
# ), f"\nExpected: {file_transcription.text}\nReceived: {streaming_transcription.text}"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|