Add Dart API for homophone replacer (#2167)

dart-api-examples/non-streaming-asr/bin/sense-voice-with-hr.dart

@@ -0,0 +1,78 @@
+// Copyright (c)  2025  Xiaomi Corporation
+import 'dart:io';
+
+import 'package:args/args.dart';
+import 'package:sherpa_onnx/sherpa_onnx.dart' as sherpa_onnx;
+
+import './init.dart';
+
+void main(List<String> arguments) async {
+  await initSherpaOnnx();
+
+  final parser = ArgParser()
+    ..addOption('model', help: 'Path to the SenseVoice model')
+    ..addOption('tokens', help: 'Path to tokens.txt')
+    ..addOption('language',
+        help: 'auto, zh, en, ja, ko, yue, or leave it empty to use auto',
+        defaultsTo: '')
+    ..addOption('use-itn',
+        help: 'true to use inverse text normalization', defaultsTo: 'false')
+    ..addOption('input-wav', help: 'Path to input.wav to transcribe')
+    ..addOption('hr-dict-dir',
+        help: 'Path to jieba dict for homophone replacer')
+    ..addOption('hr-lexicon',
+        help: 'Path to lexicon.txt for homophone replacer')
+    ..addOption('hr-rule-fsts',
+        help: 'Path to replace.fst for homophone replacer');
+
+  final res = parser.parse(arguments);
+  if (res['model'] == null ||
+      res['tokens'] == null ||
+      res['hr-dict-dir'] == null ||
+      res['hr-lexicon'] == null ||
+      res['hr-rule-fsts'] == null ||
+      res['input-wav'] == null) {
+    print(parser.usage);
+    exit(1);
+  }
+
+  final model = res['model'] as String;
+  final tokens = res['tokens'] as String;
+  final inputWav = res['input-wav'] as String;
+  final language = res['language'] as String;
+  final useItn = (res['use-itn'] as String).toLowerCase() == 'true';
+  final hrDictDir = res['hr-dict-dir'] as String;
+  final hrLexicon = res['hr-lexicon'] as String;
+  final hrRuleFsts = res['hr-rule-fsts'] as String;
+
+  final senseVoice = sherpa_onnx.OfflineSenseVoiceModelConfig(
+      model: model, language: language, useInverseTextNormalization: useItn);
+
+  final modelConfig = sherpa_onnx.OfflineModelConfig(
+    senseVoice: senseVoice,
+    tokens: tokens,
+    debug: true,
+    numThreads: 1,
+  );
+
+  final hr = sherpa_onnx.HomophoneReplacerConfig(
+      dictDir: hrDictDir, lexicon: hrLexicon, ruleFsts: hrRuleFsts);
+
+  final config =
+      sherpa_onnx.OfflineRecognizerConfig(model: modelConfig, hr: hr);
+
+  final recognizer = sherpa_onnx.OfflineRecognizer(config);
+
+  final waveData = sherpa_onnx.readWave(inputWav);
+  final stream = recognizer.createStream();
+
+  stream.acceptWaveform(
+      samples: waveData.samples, sampleRate: waveData.sampleRate);
+  recognizer.decode(stream);
+
+  final result = recognizer.getResult(stream);
+  print(result.text);
+
+  stream.free();
+  recognizer.free();
+}

dart-api-examples/non-streaming-asr/run-sense-voice-with-hr.sh

@@ -0,0 +1,31 @@
+#!/usr/bin/env bash
+
+set -ex
+
+dart pub get
+
+if [ ! -f ./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/tokens.txt ]; then
+  curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
+  tar xvf sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
+  rm sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
+fi
+
+if [ ! -d dict ]; then
+  curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/dict.tar.bz2
+  tar xf dict.tar.bz2
+  rm dict.tar.bz2
+
+  curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/replace.fst
+  curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/test-hr.wav
+  curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/lexicon.txt
+fi
+
+dart run \
+  ./bin/sense-voice-with-hr.dart \
+  --model ./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/model.int8.onnx \
+  --tokens ./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/tokens.txt \
+  --use-itn true \
+  --hr-dict-dir ./dict \
+  --hr-lexicon ./lexicon.txt \
+  --hr-rule-fsts ./replace.fst \
+  --input-wav ./test-hr.wav