def test_bad_recognition_response(self, mock_translate_hook,
mock_speech_hook):
mock_speech_hook.return_value.recognize_speech.return_value = RecognizeResponse(
results=[SpeechRecognitionResult()])
op = GcpTranslateSpeechOperator(
audio={"uri": "gs://bucket/object"},
config={"encoding": "LINEAR16"},
target_language='pl',
format_='text',
source_language=None,
model='base',
gcp_conn_id=GCP_CONN_ID,
task_id='id',
)
with self.assertRaises(AirflowException) as cm:
op.execute(context=None)
err = cm.exception
self.assertIn("it should contain 'alternatives' field", str(err))
mock_speech_hook.assert_called_once_with(gcp_conn_id=GCP_CONN_ID)
mock_translate_hook.assert_called_once_with(gcp_conn_id=GCP_CONN_ID)
mock_speech_hook.return_value.recognize_speech.assert_called_once_with(
audio={"uri": "gs://bucket/object"},
config={"encoding": "LINEAR16"},
)
mock_translate_hook.return_value.translate.assert_not_called()
def test_minimal_green_path(self, mock_translate_hook, mock_speech_hook):
mock_speech_hook.return_value.recognize_speech.return_value = RecognizeResponse(
results=[
SpeechRecognitionResult(alternatives=[
SpeechRecognitionAlternative(
transcript='test speech recognition result')
])
])
mock_translate_hook.return_value.translate.return_value = [{
'translatedText':
'sprawdzić wynik rozpoznawania mowy',
'detectedSourceLanguage':
'en',
'model':
'base',
'input':
'test speech recognition result',
}]
op = GcpTranslateSpeechOperator(
audio={"uri": "gs://bucket/object"},
config={"encoding": "LINEAR16"},
target_language='pl',
format_='text',
source_language=None,
model='base',
gcp_conn_id=GCP_CONN_ID,
task_id='id',
)
return_value = op.execute(context=None)
mock_speech_hook.assert_called_once_with(gcp_conn_id=GCP_CONN_ID)
mock_translate_hook.assert_called_once_with(gcp_conn_id=GCP_CONN_ID)
mock_speech_hook.return_value.recognize_speech.assert_called_once_with(
audio={"uri": "gs://bucket/object"},
config={"encoding": "LINEAR16"},
)
mock_translate_hook.return_value.translate.assert_called_once_with(
values='test speech recognition result',
target_language='pl',
format_='text',
source_language=None,
model='base',
)
self.assertEqual(
[{
'translatedText': 'sprawdzić wynik rozpoznawania mowy',
'detectedSourceLanguage': 'en',
'model': 'base',
'input': 'test speech recognition result',
}],
return_value,
)
voice=VOICE,
audio_config=AUDIO_CONFIG,
target_bucket_name=BUCKET_NAME,
target_filename=FILENAME,
task_id="text_to_speech_synthesize_task",
)
# [END howto_operator_text_to_speech_synthesize]
# [START howto_operator_speech_to_text_recognize]
speech_to_text_recognize_task = CloudSpeechToTextRecognizeSpeechOperator(
project_id=GCP_PROJECT_ID, config=CONFIG, audio=AUDIO, task_id="speech_to_text_recognize_task"
)
# [END howto_operator_speech_to_text_recognize]
text_to_speech_synthesize_task >> speech_to_text_recognize_task
# [START howto_operator_translate_speech]
translate_speech_task = GcpTranslateSpeechOperator(
project_id=GCP_PROJECT_ID,
audio=AUDIO,
config=CONFIG,
target_language=TARGET_LANGUAGE,
format_=FORMAT,
source_language=SOURCE_LANGUAGE,
model=MODEL,
task_id='translate_speech_task'
)
# [END howto_operator_translate_speech]
text_to_speech_synthesize_task >> translate_speech_task