def test_can_transcribe_long_speech(self):
manager = VideoManager(
os.path.join(os.path.dirname(__file__), "input", "test1.mp4"), )
_, audio = tempfile.mkstemp(".wav")
manager.extract_audio(audio)
audio_manager = AudioManager(audio)
transcriber = Transcriber(audio_manager)
transcribe_result = list(transcriber.transcribe())
self.assertGreaterEqual(len(transcribe_result), 1)
def test_can_extract_audio(self):
manager = VideoManager(
os.path.join(os.path.dirname(__file__), "input", "test1.mp4"),
)
_, audio = tempfile.mkstemp(".wav")
extracted_audio = manager.extract_audio(audio)
self.assertEqual(audio, extracted_audio)
self.assertTrue(os.path.exists(extracted_audio))
os.remove(audio)
def main():
logger = logging.getLogger()
logger.info("start processing %s", input_file)
manager = VideoManager(input_file)
audio = manager.extract_audio(settings.ROOT / "output/a.wav")
thumbnail = manager.extract_thumbnail(settings.ROOT / "output/a.jpg")
audio_manager = AudioManager(audio)
transcriber = Transcriber(audio_manager)
filter_sections = []
for transcribtion in transcriber.transcribe():
logger.debug("transcription: %s", transcribtion)
word, start, end = transcribtion
if detect_mature_word(word):
logger.debug("mature word: %s, %s", word, detect_mature_word(word))
audio_manager.apply_beep(start, end)
manager.apply_mask(start, end)
filter_sections.append({
"start_time": start,
"end_time": end,
"word": word
})
manager.apply_audio(audio_manager.save(settings.ROOT /
"output/a_beep.wav"))
manager.save(settings.ROOT / "output/a.mp4")
print(
json.dumps({
"thumbnail": str(thumbnail),
"filter_sections": filter_sections,
"filter_video": str(settings.ROOT / "output" / "a.mp4"),
}))