Support passing utf-8 strings from JavaScript to C++.

We first convert utf-16 strings to Uint8Array and then we pass the array to C++.
k2-fsa · Sep 18, 2024 · 2be8d58 · 2be8d58
1 parent bf06b26
commit 2be8d58
Show file tree

Hide file tree

Showing 3 changed files with 34 additions and 12 deletions.
diff --git a/nodejs-addon-examples/package.json b/nodejs-addon-examples/package.json
@@ -1,5 +1,5 @@
 {
   "dependencies": {
-    "sherpa-onnx-node": "^1.10.26"
+    "sherpa-onnx-node": "^1.10.27"
   }
 }
diff --git a/nodejs-addon-examples/test_asr_non_streaming_sense_voice.js b/nodejs-addon-examples/test_asr_non_streaming_sense_voice.js
@@ -3,6 +3,19 @@ const sherpa_onnx = require('sherpa-onnx-node');
 
 // Please download test files from
 // https://github.com/k2-fsa/sherpa-onnx/releases/tag/asr-models
+
+
+// If your path contains non-ascii characters, e.g., Chinese, you can use
+// the following code
+//
+
+// let encoder = new TextEncoder();
+// let tokens = encoder.encode(
+//     './sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/测试.txt');
+// let model = encoder.encode(
+//     './sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/测试.int8.onnx');
+
+
 const config = {
   'featConfig': {
     'sampleRate': 16000,
@@ -12,9 +25,11 @@ const config = {
     'senseVoice': {
       'model':
           './sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/model.int8.onnx',
+      // 'model': model,
       'useInverseTextNormalization': 1,
     },
     'tokens': './sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/tokens.txt',
+    // 'tokens': tokens,
     'numThreads': 2,
     'provider': 'cpu',
     'debug': 1,

diff --git a/scripts/node-addon-api/src/macros.h b/scripts/node-addon-api/src/macros.h
@@ -7,17 +7,24 @@
 #include <algorithm>
 #include <string>
 
-#define SHERPA_ONNX_ASSIGN_ATTR_STR(c_name, js_name)          \
-  do {                                                        \
-    if (o.Has(#js_name) && o.Get(#js_name).IsString()) {      \
-      Napi::String _str = o.Get(#js_name).As<Napi::String>(); \
-      std::string s = _str.Utf8Value();                       \
-      char *p = new char[s.size() + 1];                       \
-      std::copy(s.begin(), s.end(), p);                       \
-      p[s.size()] = 0;                                        \
-                                                              \
-      c.c_name = p;                                           \
-    }                                                         \
+#define SHERPA_ONNX_ASSIGN_ATTR_STR(c_name, js_name)                       \
+  do {                                                                     \
+    if (o.Has(#js_name) && o.Get(#js_name).IsString()) {                   \
+      Napi::String _str = o.Get(#js_name).As<Napi::String>();              \
+      std::string s = _str.Utf8Value();                                    \
+      char *p = new char[s.size() + 1];                                    \
+      std::copy(s.begin(), s.end(), p);                                    \
+      p[s.size()] = 0;                                                     \
+                                                                           \
+      c.c_name = p;                                                        \
+    } else if (o.Has(#js_name) && o.Get(#js_name).IsTypedArray()) {        \
+      Napi::Uint8Array _array = o.Get(#js_name).As<Napi::Uint8Array>();    \
+      char *p = new char[_array.ElementLength() + 1];                      \
+      std::copy(_array.Data(), _array.Data() + _array.ElementLength(), p); \
+      p[_array.ElementLength()] = '\0';                                    \
+                                                                           \
+      c.c_name = p;                                                        \
+    }                                                                      \
   } while (0)
 
 #define SHERPA_ONNX_ASSIGN_ATTR_INT32(c_name, js_name)            \