def _process(self, input_file, language_code):
if not input_file.endswith('.flac'):
raise RuntimeError('Only flac encoding file is supported.')
audio_content = cloud_speech.RecognitionAudio(content=open(
input_file, 'rb').read(), )
sample_rate = self._sample_rate(input_file)
operation = self.service.AsyncRecognize(
cloud_speech.AsyncRecognizeRequest(
config=cloud_speech.RecognitionConfig(
encoding=AUDIO_ENCODING,
sample_rate=sample_rate,
language_code=language_code,
),
audio=audio_content), DEADLINE_SECS)
return operation
def main(input_uri, encoding, sample_rate):
channel = make_channel('speech.googleapis.com', 443)
service = cloud_speech_pb2.beta_create_Speech_stub(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
response = 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/best-practices#language_support
# for a list of supported languages.
language_code='fi-FI', # a BCP-47 language tag
),
audio=cloud_speech_pb2.RecognitionAudio(
uri=input_uri,
)
), DEADLINE_SECS)
# Print the longrunning operation handle.
print >> sys.stderr, response
# Construct a long running operation endpoint.
service = operations_grpc_pb2.beta_create_Operations_stub(channel)
name = response.name
while True:
# Give the server a few seconds to process.
print >> sys.stderr, 'Waiting for server processing...'
time.sleep(1)
# Get the long running operation with response.
response = service.GetOperation(
operations_grpc_pb2.GetOperationRequest(name=name),
DEADLINE_SECS)
if response.done:
break
# Print the recognition results.
results = cloud_speech_pb2.AsyncRecognizeResponse()
response.response.Unpack(results)
for result in results.results:
for alternative in result.alternatives:
print(('"{}",{}').format(alternative.transcript.encode('utf-8'), alternative.confidence))
def main(input_uri, encoding, sample_rate, language_code='en-US'):
channel = make_channel('speech.googleapis.com', 443)
service = cloud_speech_pb2.beta_create_Speech_stub(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_grpc_pb2.beta_create_Operations_stub(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_grpc_pb2.GetOperationRequest(name=name), DEADLINE_SECS)
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))
def main(input_uri, encoding, sample_rate):
channel = make_channel('speech.googleapis.com', 443)
service = cloud_speech_pb2.beta_create_Speech_stub(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
response = service.AsyncRecognize(
cloud_speech_pb2.AsyncRecognizeRequest(
config=cloud_speech_pb2.RecognitionConfig(
encoding=encoding,
sample_rate=sample_rate,
),
audio=cloud_speech_pb2.RecognitionAudio(uri=input_uri, )),
DEADLINE_SECS)
# Print the longrunning operation handle.
print(response)
# Construct a long running operation endpoint.
service = operations_grpc_pb2.beta_create_Operations_stub(channel)
name = response.name
while True:
# Give the server a few seconds to process.
print('Waiting for server processing...')
time.sleep(1)
# Get the long running operation with response.
response = service.GetOperation(
operations_grpc_pb2.GetOperationRequest(name=name), DEADLINE_SECS)
if response.done:
break
# Print the recognition results.
results = cloud_speech_pb2.AsyncRecognizeResponse()
response.response.Unpack(results)
print(results)