Add C API for Kokoro TTS 1.0 (#1801)

.github/workflows/c-api.yaml

@@ -100,6 +100,27 @@ jobs:
           rm ./kws-c-api
           rm -rf sherpa-onnx-kws-*
 
+      - name: Test Kokoro TTS (zh+en)
+        shell: bash
+        run: |
+          gcc -o kokoro-tts-zh-en-c-api ./c-api-examples/kokoro-tts-zh-en-c-api.c \
+            -I ./build/install/include \
+            -L ./build/install/lib/ \
+            -l sherpa-onnx-c-api \
+            -l onnxruntime
+
+          curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/kokoro-multi-lang-v1_0.tar.bz2
+          tar xf kokoro-multi-lang-v1_0.tar.bz2
+          rm kokoro-multi-lang-v1_0.tar.bz2
+
+          export LD_LIBRARY_PATH=$PWD/build/install/lib:$LD_LIBRARY_PATH
+          export DYLD_LIBRARY_PATH=$PWD/build/install/lib:$DYLD_LIBRARY_PATH
+
+          ./kokoro-tts-zh-en-c-api
+
+          rm ./kokoro-tts-zh-en-c-api
+          rm -rf kokoro-zh-en-*
+
       - name: Test Kokoro TTS (en)
         shell: bash
         run: |

c-api-examples/CMakeLists.txt

@@ -19,6 +19,9 @@ if(SHERPA_ONNX_ENABLE_TTS)
 
   add_executable(kokoro-tts-en-c-api kokoro-tts-en-c-api.c)
   target_link_libraries(kokoro-tts-en-c-api sherpa-onnx-c-api)
+
+  add_executable(kokoro-tts-zh-en-c-api kokoro-tts-zh-en-c-api.c)
+  target_link_libraries(kokoro-tts-zh-en-c-api sherpa-onnx-c-api)
 endif()
 
 if(SHERPA_ONNX_ENABLE_SPEAKER_DIARIZATION)

c-api-examples/kokoro-tts-zh-en-c-api.c

@@ -0,0 +1,82 @@
+// c-api-examples/kokoro-tts-zh-en-c-api.c
+//
+// Copyright (c)  2025  Xiaomi Corporation
+
+// This file shows how to use sherpa-onnx C API
+// for English + Chinese TTS with Kokoro.
+//
+// clang-format off
+/*
+Usage
+
+
+wget https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/kokoro-multi-lang-v1_0.tar.bz2
+tar xf kokoro-multi-lang-v1_0.tar.bz2
+rm kokoro-multi-lang-v1_0.tar.bz2
+
+./kokoro-tts-zh-en-c-api
+
+ */
+// clang-format on
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "sherpa-onnx/c-api/c-api.h"
+
+static int32_t ProgressCallback(const float *samples, int32_t num_samples,
+                                float progress) {
+  fprintf(stderr, "Progress: %.3f%%\n", progress * 100);
+  // return 1 to continue generating
+  // return 0 to stop generating
+  return 1;
+}
+
+int32_t main(int32_t argc, char *argv[]) {
+  SherpaOnnxOfflineTtsConfig config;
+  memset(&config, 0, sizeof(config));
+  config.model.kokoro.model = "./kokoro-multi-lang-v1_0/model.onnx";
+  config.model.kokoro.voices = "./kokoro-multi-lang-v1_0/voices.bin";
+  config.model.kokoro.tokens = "./kokoro-multi-lang-v1_0/tokens.txt";
+  config.model.kokoro.data_dir = "./kokoro-multi-lang-v1_0/espeak-ng-data";
+  config.model.kokoro.dict_dir = "./kokoro-multi-lang-v1_0/dict";
+  config.model.kokoro.lexicon =
+      "./kokoro-multi-lang-v1_0/lexicon-us-en.txt,./kokoro-multi-lang-v1_0/"
+      "lexicon-zh.txt";
+
+  config.model.num_threads = 2;
+
+  // If you don't want to see debug messages, please set it to 0
+  config.model.debug = 1;
+
+  const char *filename = "./generated-kokoro-zh-en.wav";
+  const char *text =
+      "中英文语音合成测试。This is generated by next generation Kaldi using "
+      "Kokoro without Misaki. 你觉得中英文说的如何呢？";
+
+  const SherpaOnnxOfflineTts *tts = SherpaOnnxCreateOfflineTts(&config);
+  int32_t sid = 0;    // there are 53 speakers
+  float speed = 1.0;  // larger -> faster in speech speed
+
+#if 0
+  // If you don't want to use a callback, then please enable this branch
+  const SherpaOnnxGeneratedAudio *audio =
+      SherpaOnnxOfflineTtsGenerate(tts, text, sid, speed);
+#else
+  const SherpaOnnxGeneratedAudio *audio =
+      SherpaOnnxOfflineTtsGenerateWithProgressCallback(tts, text, sid, speed,
+                                                       ProgressCallback);
+#endif
+
+  SherpaOnnxWriteWave(audio->samples, audio->n, audio->sample_rate, filename);
+
+  SherpaOnnxDestroyOfflineTtsGeneratedAudio(audio);
+  SherpaOnnxDestroyOfflineTts(tts);
+
+  fprintf(stderr, "Input text is: %s\n", text);
+  fprintf(stderr, "Speaker ID is is: %d\n", sid);
+  fprintf(stderr, "Saved to: %s\n", filename);
+
+  return 0;
+}

sherpa-onnx/c-api/c-api.cc

@@ -1120,6 +1120,10 @@ static sherpa_onnx::OfflineTtsConfig GetOfflineTtsConfig(
       SHERPA_ONNX_OR(config->model.kokoro.data_dir, "");
   tts_config.model.kokoro.length_scale =
       SHERPA_ONNX_OR(config->model.kokoro.length_scale, 1.0);
+  tts_config.model.kokoro.dict_dir =
+      SHERPA_ONNX_OR(config->model.kokoro.dict_dir, "");
+  tts_config.model.kokoro.lexicon =
+      SHERPA_ONNX_OR(config->model.kokoro.lexicon, "");
 
   tts_config.model.num_threads = SHERPA_ONNX_OR(config->model.num_threads, 1);
   tts_config.model.debug = config->model.debug;

sherpa-onnx/c-api/c-api.h

@@ -926,6 +926,8 @@ SHERPA_ONNX_API typedef struct SherpaOnnxOfflineTtsKokoroModelConfig {
   const char *data_dir;
 
   float length_scale;  // < 1, faster in speech speed; > 1, slower in speed
+  const char *dict_dir;
+  const char *lexicon;
 } SherpaOnnxOfflineTtsKokoroModelConfig;
 
 SHERPA_ONNX_API typedef struct SherpaOnnxOfflineTtsModelConfig {

sherpa-onnx/csrc/offline-tts-kokoro-impl.h

@@ -4,6 +4,8 @@
 #ifndef SHERPA_ONNX_CSRC_OFFLINE_TTS_KOKORO_IMPL_H_
 #define SHERPA_ONNX_CSRC_OFFLINE_TTS_KOKORO_IMPL_H_
 
+#include <iomanip>
+#include <ios>
 #include <memory>
 #include <string>
 #include <strstream>
@@ -188,6 +190,20 @@ class OfflineTtsKokoroImpl : public OfflineTtsImpl {
       SHERPA_ONNX_LOGE("Raw text: %{public}s", text.c_str());
 #else
       SHERPA_ONNX_LOGE("Raw text: %s", text.c_str());
+#endif
+      std::ostringstream os;
+      os << "In bytes (hex):\n";
+      const auto p = reinterpret_cast<const uint8_t *>(text.c_str());
+      for (int32_t i = 0; i != text.size(); ++i) {
+        os << std::setw(2) << std::setfill('0') << std::hex
+           << static_cast<uint32_t>(p[i]) << " ";
+      }
+      os << "\n";
+
+#if __OHOS__
+      SHERPA_ONNX_LOGE("%{public}s", os.str().c_str());
+#else
+      SHERPA_ONNX_LOGE("%s", os.str().c_str());
 #endif
     }