From 6631b9fead2c60ac2a52fe9407657ef4cec6fabb Mon Sep 17 00:00:00 2001
From: Jarno Hakulinen <jhakulin@microsoft.com>
Date: Tue, 6 Dec 2022 11:21:16 -0800
Subject: [PATCH] Add AudioStreamWaveFormat with ALAW, MULAW support (#82)

* add AudioStreamWaveFormat with alaw, mulaw support

* fix lint
---
 audio/audio_stream_container_format.go | 15 +++++++
 audio/audio_stream_format.go           | 17 ++++++++
 samples/main.go                        |  1 +
 samples/recognizer/from_file.go        | 58 ++++++++++++++++++++++++++
 4 files changed, 91 insertions(+)

diff --git a/audio/audio_stream_container_format.go b/audio/audio_stream_container_format.go
index fea3fa6..97823bf 100644
--- a/audio/audio_stream_container_format.go
+++ b/audio/audio_stream_container_format.go
@@ -31,3 +31,18 @@ const (
 	// ANY Stream ContainerFormat definition when the actual stream format is not known.
 	ANY AudioStreamContainerFormat = 0x108
 )
+
+// AudioStreamWaveFormat represents the format specified inside WAV container which are sent directly as encoded to the speech service.
+type AudioStreamWaveFormat int //nolint:revive
+
+const (
+	// AudioStreamWaveFormat definition for PCM (pulse-code modulated) data in integer format.
+	WavePCM AudioStreamWaveFormat = 0x0001
+
+	// AudioStreamWaveFormat definition A-law-encoded format.
+	WaveALAW AudioStreamWaveFormat = 0x0006
+
+	// AudioStreamWaveFormat definition for Mu-law-encoded format.
+	WaveMULAW AudioStreamWaveFormat = 0x0007
+
+)
diff --git a/audio/audio_stream_format.go b/audio/audio_stream_format.go
index 3c5c81e..d8767a0 100644
--- a/audio/audio_stream_format.go
+++ b/audio/audio_stream_format.go
@@ -32,6 +32,23 @@ func GetDefaultInputFormat() (*AudioStreamFormat, error) {
 	return format, nil
 }
 
+// GetWaveFormat creates an audio stream format object with the specified waveformat characteristics.
+func GetWaveFormat(samplesPerSecond uint32, bitsPerSample uint8, channels uint8, waveFormat AudioStreamWaveFormat) (*AudioStreamFormat, error) {
+	var handle C.SPXHANDLE
+	ret := uintptr(C.audio_stream_format_create_from_waveformat(
+		&handle,
+		(C.uint32_t)(samplesPerSecond),
+		(C.uint8_t)(bitsPerSample),
+		(C.uint8_t)(channels),
+		(C.Audio_Stream_Wave_Format)(waveFormat)))
+	if ret != C.SPX_NOERROR {
+		return nil, common.NewCarbonError(ret)
+	}
+	format := new(AudioStreamFormat)
+	format.handle = handle
+	return format, nil
+}
+
 // GetWaveFormatPCM creates an audio stream format object with the specified PCM waveformat characteristics.
 // Note: Currently, only WAV / PCM with 16-bit samples, 16 kHz sample rate, and a single channel (Mono) is supported. When
 // used with Conversation Transcription, eight channels are supported.
diff --git a/samples/main.go b/samples/main.go
index bf61a44..f093513 100644
--- a/samples/main.go
+++ b/samples/main.go
@@ -31,6 +31,7 @@ func main() {
 	samples := functionMap{
 		"speech_recognizer:RecognizeOnceFromWavFile":        recognizer.RecognizeOnceFromWavFile,
 		"speech_recognizer:RecognizeOnceFromCompressedFile": recognizer.RecognizeOnceFromCompressedFile,
+		"speech_recognizer:RecognizeOnceFromALAWFile":       recognizer.RecognizeOnceFromALAWFile,
 		"speech_recognizer:ContinuousFromMicrophone":        recognizer.ContinuousFromMicrophone,
 		"speech_recognizer:RecognizeContinuousUsingWrapper": recognizer.RecognizeContinuousUsingWrapper,
 		"dialog_service_connector:ListenOnce":               dialog_service_connector.ListenOnce,
diff --git a/samples/recognizer/from_file.go b/samples/recognizer/from_file.go
index cf25d3d..c9c9bde 100644
--- a/samples/recognizer/from_file.go
+++ b/samples/recognizer/from_file.go
@@ -125,3 +125,61 @@ func RecognizeOnceFromCompressedFile(subscription string, region string, file st
 	fmt.Println("Got a recognition!")
 	fmt.Println(outcome.Result.Text)
 }
+
+func RecognizeOnceFromALAWFile(subscription string, region string, file string) {
+	var waveFormat audio.AudioStreamWaveFormat
+	waveFormat = audio.WaveALAW
+	format, err := audio.GetWaveFormat(8000, 16, 1, waveFormat)
+	if err != nil {
+		fmt.Println("Got an error: ", err)
+		return
+	}
+	defer format.Close()
+	stream, err := audio.CreatePushAudioInputStreamFromFormat(format)
+	if err != nil {
+		fmt.Println("Got an error: ", err)
+		return
+	}
+	defer stream.Close()
+	audioConfig, err := audio.NewAudioConfigFromStreamInput(stream)
+	if err != nil {
+		fmt.Println("Got an error: ", err)
+		return
+	}
+	defer audioConfig.Close()
+	config, err := speech.NewSpeechConfigFromSubscription(subscription, region)
+	if err != nil {
+		fmt.Println("Got an error: ", err)
+		return
+	}
+	defer config.Close()
+	speechRecognizer, err := speech.NewSpeechRecognizerFromConfig(config, audioConfig)
+	if err != nil {
+		fmt.Println("Got an error: ", err)
+		return
+	}
+	defer speechRecognizer.Close()
+	speechRecognizer.SessionStarted(func(event speech.SessionEventArgs) {
+		defer event.Close()
+		fmt.Println("Session Started (ID=", event.SessionID, ")")
+	})
+	speechRecognizer.SessionStopped(func(event speech.SessionEventArgs) {
+		defer event.Close()
+		fmt.Println("Session Stopped (ID=", event.SessionID, ")")
+	})
+	helpers.PumpFileIntoStream(file, stream)
+	task := speechRecognizer.RecognizeOnceAsync()
+	var outcome speech.SpeechRecognitionOutcome
+	select {
+	case outcome = <-task:
+	case <-time.After(40 * time.Second):
+		fmt.Println("Timed out")
+		return
+	}
+	defer outcome.Close()
+	if outcome.Error != nil {
+		fmt.Println("Got an error: ", outcome.Error)
+	}
+	fmt.Println("Got a recognition!")
+	fmt.Println(outcome.Result.Text)
+}
\ No newline at end of file