Open-source AI podcast clipper. Generate vertical clips with face tracking and burned-in captions. CLI, MCP server, web app.
Python ai-toolsauto-captionscaptionscliffmpegmcpmcp-serveropusclip-alternativepodcast-clipspodcastsshortform-videovertical-videowhisper
import json
import os
import sys
import tempfile
import unittest
from unittest import mock
ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
BACKEND_ROOT = os.path.join(ROOT, "backend")
if BACKEND_ROOT not in sys.path:
sys.path.insert(0, BACKEND_ROOT)
from services.transcription_whispercpp import (
_dtw_preset_for_model,
_tokens_to_words,
transcribe_file,
)
class WhisperCppAdapterTests(unittest.TestCase):
def test_sentencepiece_marker_is_removed(self):
words = _tokens_to_words([
{"text": "▁hello", "offsets": {"from": 0, "to": 100}},
{"text": "▁world", "offsets": {"from": 100, "to": 200}},
])
self.assertEqual([w["word"] for w in words], ["hello", "world"])
class LanguageHandlingTests(unittest.TestCase):
"""whisper-cli defaults to English decoding when -l is omitted entirely, so
an unset language must become "-l auto" rather than no flag at all; the
reported language must come from whisper-cli's detected result, not the
(possibly "auto") param we passed in."""
def _fake_run(self, result_language, params_language):
def run(cmd, **kwargs):
of_index = cmd.index("-of")
out_base = cmd[of_index + 1]
payload = {
"params": {"language": params_language},
"result": {"language": result_language},
"transcription": [],
}
with open(out_base + ".json", "w", encoding="utf-8") as f:
json.dump(payload, f)
return mock.Mock(returncode=0)
return mock.Mock(side_effect=run)
def _run_transcribe(self, language, result_language, params_language):
with tempfile.NamedTemporaryFile(suffix=".mp3") as media, \
tempfile.NamedTemporaryFile(suffix=".bin") as model, \
tempfile.NamedTemporaryFile(suffix=".wav") as wav:
fake_run = self._fake_run(result_language, params_language)
with mock.patch("services.transcription_whispercpp.subprocess.run", fake_run):
result = transcribe_file(
media.name,
model.name,
language=language,
wav_path=wav.name,
)
cmd = fake_run.call_args[0][0]
return result, cmd
def test_unset_language_passes_auto(self):
_, cmd = self._run_transcribe(None, "es", "auto")
self.assertIn("-l", cmd)
self.assertEqual(cmd[cmd.index("-l") + 1], "auto")
def test_explicit_language_is_passed_through(self):
_, cmd = self._run_transcribe("fr", "fr", "fr")
self.assertEqual(cmd[cmd.index("-l") + 1], "fr")
def test_label_prefers_detected_result_language_over_param(self):
# Requesting auto-detect but whisper-cli actually detects Spanish: the
# output must be labeled "es", never the "auto" we passed as -l.
result, _ = self._run_transcribe(None, "es", "auto")
self.assertEqual(result["language"], "es")
class DtwPresetTests(unittest.TestCase):
def test_preset_tracks_the_model_file(self):
cases = {
"ggml-tiny.en.bin": "tiny.en",
"ggml-base.bin": "base",
"ggml-small.bin": "small",
"ggml-large-v3-turbo.bin": "large.v3-turbo",
"ggml-base.en-q5_1.bin": "base.en",
# K-quantisation names carry underscores between the qualifiers, and
# failing to strip them dropped -dtw for a model that supports it.
"ggml-large-v3-q4_k_m.gguf": "large.v3",
"ggml-small.en-q8_0.bin": "small.en",
}
for name, preset in cases.items():
with self.subTest(name=name):
self.assertEqual(_dtw_preset_for_model(os.path.join("/models", name)), preset)
def test_unknown_model_gets_no_preset(self):
self.assertIsNone(_dtw_preset_for_model("/models/ggml-distil-large-v2.bin"))
if __name__ == "__main__":
unittest.main()