def async_recognize(self, config, audio, options=None):
"""
Perform asynchronous speech-recognition: receive results via the
google.longrunning.Operations interface. Returns either an
``Operation.error`` or an ``Operation.response`` which contains
an ``AsyncRecognizeResponse`` message.
Example:
>>> from google.cloud.gapic.speech.v1beta1 import speech_client
>>> from google.cloud.grpc.speech.v1beta1 import cloud_speech_pb2
>>> api = speech_client.SpeechClient()
>>> config = cloud_speech_pb2.RecognitionConfig()
>>> audio = cloud_speech_pb2.RecognitionAudio()
>>> response = api.async_recognize(config, audio)
Args:
config (:class:`google.cloud.grpc.speech.v1beta1.cloud_speech_pb2.RecognitionConfig`): [Required] The ``config`` message provides information to the recognizer
that specifies how to process the request.
audio (:class:`google.cloud.grpc.speech.v1beta1.cloud_speech_pb2.RecognitionAudio`): [Required] The audio data to be recognized.
options (:class:`google.gax.CallOptions`): Overrides the default
settings for this call, e.g, timeout, retries etc.
Returns:
A :class:`google.longrunning.operations_pb2.Operation` instance.
Raises:
:exc:`google.gax.errors.GaxError` if the RPC is aborted.
:exc:`ValueError` if the parameters are invalid.
"""
request = cloud_speech_pb2.AsyncRecognizeRequest(
config=config, audio=audio)
return self._async_recognize(request, options)
def main(input_uri, encoding, sample_rate, language_code='en-US'):
channel = make_channel('speech.googleapis.com', 443)
service = cloud_speech_pb2.SpeechStub(channel)
# The method and parameters can be inferred from the proto from which the
# grpc client lib was generated. See:
# https://github.com/googleapis/googleapis/blob/master/google/cloud/speech/v1beta1/cloud_speech.proto
operation = service.AsyncRecognize(
cloud_speech_pb2.AsyncRecognizeRequest(
config=cloud_speech_pb2.RecognitionConfig(
# There are a bunch of config options you can specify. See
# https://goo.gl/KPZn97 for the full list.
encoding=encoding, # one of LINEAR16, FLAC, MULAW, AMR, AMR_WB
sample_rate=sample_rate, # the rate in hertz
# See https://g.co/cloud/speech/docs/languages for a list of
# supported languages.
language_code=language_code, # a BCP-47 language tag
),
audio=cloud_speech_pb2.RecognitionAudio(uri=input_uri, )),
DEADLINE_SECS)
# Print the longrunning operation handle.
print(operation)
# Construct a long running operation endpoint.
service = operations_pb2.OperationsStub(channel)
name = operation.name
while True:
# Give the server a few seconds to process.
print('Waiting for server processing...')
time.sleep(1)
operation = service.GetOperation(
operations_pb2.GetOperationRequest(name=name), DEADLINE_SECS)
if operation.error.message:
print('\nOperation error:\n{}'.format(operation.error))
if operation.done:
break
response = cloud_speech_pb2.AsyncRecognizeResponse()
operation.response.Unpack(response)
# Print the recognition result alternatives and confidence scores.
for result in response.results:
print('Result:')
for alternative in result.alternatives:
print(u' ({}): {}'.format(alternative.confidence,
alternative.transcript))