EngineX-Iluvatar/enginex_bi_series-sherpa-onnx
Archived
15
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

update kotlin api for better release native object and add user-friendly apis. (#1275)

Browse Source
This commit is contained in:
Robin Zhong
2024-08-22 19:18:11 +08:00
committed by GitHub
parent 5a2aa110b8
commit d8001d6edc
8 changed files with 100 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
sherpa-onnx/kotlin-api/KeywordSpotter.kt
View File

@@ -24,7 +24,7 @@ class KeywordSpotter(
assetManager: AssetManager? = null,
val config: KeywordSpotterConfig,
) {
private val ptr: Long
private var ptr: Long
init {
ptr = if (assetManager != null) {
@@ -35,7 +35,10 @@ class KeywordSpotter(
}
protected fun finalize() {
delete(ptr)
if (ptr != 0L) {
delete(ptr)
ptr = 0
}
}
fun release() = finalize()

7
sherpa-onnx/kotlin-api/OfflinePunctuation.kt
View File

@@ -18,7 +18,7 @@ class OfflinePunctuation(
assetManager: AssetManager? = null,
config: OfflinePunctuationConfig,
) {
private val ptr: Long
private var ptr: Long
init {
ptr = if (assetManager != null) {
@@ -29,7 +29,10 @@ class OfflinePunctuation(
}
protected fun finalize() {
delete(ptr)
if (ptr != 0L) {
delete(ptr)
ptr = 0
}
}
fun release() = finalize()

16
sherpa-onnx/kotlin-api/OfflineRecognizer.kt
View File

@@ -72,7 +72,7 @@ class OfflineRecognizer(
assetManager: AssetManager? = null,
config: OfflineRecognizerConfig,
) {
private val ptr: Long
private var ptr: Long
init {
ptr = if (assetManager != null) {
@@ -83,7 +83,10 @@ class OfflineRecognizer(
}
protected fun finalize() {
delete(ptr)
if (ptr != 0L) {
delete(ptr)
ptr = 0
}
}
fun release() = finalize()
@@ -102,7 +105,14 @@ class OfflineRecognizer(
val lang = objArray[3] as String
val emotion = objArray[4] as String
val event = objArray[5] as String
return OfflineRecognizerResult(text = text, tokens = tokens, timestamps = timestamps, lang = lang, emotion = emotion, event = event)
return OfflineRecognizerResult(
text = text,
tokens = tokens,
timestamps = timestamps,
lang = lang,
emotion = emotion,
event = event
)
}
fun decode(stream: OfflineStream) = decode(ptr, stream.ptr)

8
sherpa-onnx/kotlin-api/OfflineStream.kt
View File

@@ -13,6 +13,14 @@ class OfflineStream(var ptr: Long) {
fun release() = finalize()
fun use(block: (OfflineStream) -> Unit) {
try {
block(this)
} finally {
release()
}
}
private external fun acceptWaveform(ptr: Long, samples: FloatArray, sampleRate: Int)
private external fun delete(ptr: Long)

9
sherpa-onnx/kotlin-api/OnlineRecognizer.kt
View File

@@ -62,7 +62,7 @@ data class OnlineRecognizerConfig(
var featConfig: FeatureConfig = FeatureConfig(),
var modelConfig: OnlineModelConfig,
var lmConfig: OnlineLMConfig = OnlineLMConfig(),
var ctcFstDecoderConfig : OnlineCtcFstDecoderConfig = OnlineCtcFstDecoderConfig(),
var ctcFstDecoderConfig: OnlineCtcFstDecoderConfig = OnlineCtcFstDecoderConfig(),
var endpointConfig: EndpointConfig = EndpointConfig(),
var enableEndpoint: Boolean = true,
var decodingMethod: String = "greedy_search",
@@ -85,7 +85,7 @@ class OnlineRecognizer(
assetManager: AssetManager? = null,
val config: OnlineRecognizerConfig,
) {
private val ptr: Long
private var ptr: Long
init {
ptr = if (assetManager != null) {
@@ -96,7 +96,10 @@ class OnlineRecognizer(
}
protected fun finalize() {
delete(ptr)
if (ptr != 0L) {
delete(ptr)
ptr = 0
}
}
fun release() = finalize()

9
sherpa-onnx/kotlin-api/OnlineStream.kt
View File

@@ -15,10 +15,19 @@ class OnlineStream(var ptr: Long = 0) {
fun release() = finalize()
fun use(block: (OnlineStream) -> Unit) {
try {
block(this)
} finally {
release()
}
}
private external fun acceptWaveform(ptr: Long, samples: FloatArray, sampleRate: Int)
private external fun inputFinished(ptr: Long)
private external fun delete(ptr: Long)
companion object {
init {
System.loadLibrary("sherpa-onnx-jni")

18
sherpa-onnx/kotlin-api/Vad.kt
View File

@@ -19,11 +19,13 @@ data class VadModelConfig(
var debug: Boolean = false,
)
class SpeechSegment(val start: Int, val samples: FloatArray)
class Vad(
assetManager: AssetManager? = null,
var config: VadModelConfig,
) {
private val ptr: Long
private var ptr: Long
init {
if (assetManager != null) {
@@ -34,17 +36,23 @@ class Vad(
}
protected fun finalize() {
delete(ptr)
if (ptr != 0L) {
delete(ptr)
ptr = 0
}
}
fun release() = finalize()
fun acceptWaveform(samples: FloatArray) = acceptWaveform(ptr, samples)
fun empty(): Boolean = empty(ptr)
fun pop() = pop(ptr)
// return an array containing
// [start: Int, samples: FloatArray]
fun front() = front(ptr)
fun front(): SpeechSegment {
val segment = front(ptr)
return SpeechSegment(segment[0] as Int, segment[1] as FloatArray)
}
fun clear() = clear(ptr)

41
sherpa-onnx/kotlin-api/WaveReader.kt
View File

@@ -3,8 +3,49 @@ package com.k2fsa.sherpa.onnx
import android.content.res.AssetManager
data class WaveData(
val samples: FloatArray,
val sampleRate: Int,
) {
override fun equals(other: Any?): Boolean {
if (this === other) return true
if (javaClass != other?.javaClass) return false
other as WaveData
if (!samples.contentEquals(other.samples)) return false
if (sampleRate != other.sampleRate) return false
return true
}
override fun hashCode(): Int {
var result = samples.contentHashCode()
result = 31 * result + sampleRate
return result
}
}
class WaveReader {
companion object {
fun readWave(
assetManager: AssetManager,
filename: String,
): WaveData {
return readWaveFromAsset(assetManager, filename).let {
WaveData(it[0] as FloatArray, it[1] as Int)
}
}
fun readWave(
filename: String,
): WaveData {
return readWaveFromFile(filename).let {
WaveData(it[0] as FloatArray, it[1] as Int)
}
}
// Read a mono wave file asset
// The returned array has two entries:
// - the first entry contains an 1-D float array