longRunningRecognize method

ResponseFuture<Operation> longRunningRecognize(RecognitionConfigV2 config, String audioGcsUri, { String location = 'global', })

Sends a LongRunningRecognizeRequest request to the Google Speech Api. Requires a RecognitionConfigBeta and an RecognitionAudio.

To use asynchronous speech recognition to transcribe audio longer than 60 seconds, you must have your data saved in a Google Cloud Storage bucket.

Implementation

ResponseFuture<Operation> longRunningRecognize(
    RecognitionConfigV2 config, String audioGcsUri,
    {String location = 'global'}) {
  final client = SpeechClient(_channel, options: _options);

  // transform audio to RecognitionAudio
  final recognitionAudio = BatchRecognizeFileMetadata()..uri = audioGcsUri;

  // Create the request, which transmits the necessary
  // data to the Google Api.
  final request = BatchRecognizeRequest(
    recognizer: 'projects/$projectId/locations/$location/recognizers/_',
    files: [recognitionAudio],
    config: config.toConfig(),
    recognitionOutputConfig:
        RecognitionOutputConfig(inlineResponseConfig: InlineOutputConfig()),
  );

  return client.batchRecognize(request);
}