Update keywords (#560)

Picovoice · May 15, 2023 · e6693fd · e6693fd
1 parent fae4bc4
commit e6693fd
Show file tree

Hide file tree

Showing 5 changed files with 223 additions and 125 deletions.
diff --git a/README.md b/README.md
@@ -21,7 +21,9 @@
 
 Made in Vancouver, Canada by [Picovoice](https://picovoice.ai)
 
+<!-- markdown-link-check-disable -->
 [![Twitter URL](https://img.shields.io/twitter/url?label=%40AiPicovoice&style=social&url=https%3A%2F%2Ftwitter.com%2FAiPicovoice)](https://twitter.com/AiPicovoice)
+<!-- markdown-link-check-enable -->
 [![YouTube Channel Views](https://img.shields.io/youtube/channel/views/UCAdi9sTCXLosG1XeqDwLx7w?label=YouTube&style=social)](https://www.youtube.com/channel/UCAdi9sTCXLosG1XeqDwLx7w)
 
 Picovoice is the end-to-end platform for building voice products on your terms. Unlike Alexa and Google services,

diff --git a/resources/.test/test_data.json b/resources/.test/test_data.json
@@ -3,7 +3,7 @@
     "parameters": [
       {
         "language": "ar",
-        "wakeword": "hummus",
+        "wakeword": "الحمص",
         "context_name": "simple_context_ar",
         "audio_file": "hummus-simple-context_ar.wav",
         "inference": {
@@ -65,7 +65,7 @@
       },
       {
         "language": "hi",
-        "wakeword": "namaste",
+        "wakeword": "नमस्ते",
         "context_name": "simple_context_hi",
         "audio_file": "namaste-simple-context_hi.wav",
         "inference": {
@@ -89,7 +89,7 @@
       },
       {
         "language": "ja",
-        "wakeword": "ninja",
+        "wakeword": "忍者",
         "context_name": "sumāto_shōmei",
         "audio_file": "ninja-sumāto-shōmei_ja.wav",
         "inference": {
@@ -101,7 +101,7 @@
       },
       {
         "language": "ko",
-        "wakeword": "koppulso",
+        "wakeword": "코뿔소",
         "context_name": "seumateu_jomyeong",
         "audio_file": "koppulso-seumateu-jomyeong_ko.wav",
         "inference": {
@@ -125,7 +125,7 @@
       },
       {
         "language": "pl",
-        "wakeword": "porcupine_pl",
+        "wakeword": "jeżozwierz",
         "context_name": "simple_context_pl",
         "audio_file": "porcupine-simple-context_pl.wav",
         "inference": {
@@ -149,7 +149,7 @@
       },
       {
         "language": "ru",
-        "wakeword": "pomogite",
+        "wakeword": "помогите",
         "context_name": "simple_context_ru",
         "audio_file": "pomogite-simple-context_ru.wav",
         "inference": {
@@ -161,7 +161,7 @@
       },
       {
         "language": "sv",
-        "wakeword": "porcupine_sv",
+        "wakeword": "piggsvin",
         "context_name": "simple_context_sv",
         "audio_file": "porcupine-simple-context_sv.wav",
         "inference": {
@@ -173,7 +173,7 @@
       },
       {
         "language": "vn",
-        "wakeword": "chao_chi",
+        "wakeword": "chào chị",
         "context_name": "simple_context_vn",
         "audio_file": "chao-chi-simple-context_vn.wav",
         "inference": {
@@ -185,7 +185,7 @@
       },
       {
         "language": "zh",
-        "wakeword": "porcupine_zh",
+        "wakeword": "豪猪",
         "context_name": "simple_context_zh",
         "audio_file": "porcupine-simple-context_zh.wav",
         "inference": {

diff --git a/resources/porcupine b/resources/porcupine
diff --git a/sdk/web/scripts/setup_test.js b/sdk/web/scripts/setup_test.js
@@ -1,5 +1,6 @@
 const fs = require('fs');
 const { join } = require('path');
+const { createHash } = require('crypto');
 
 console.log('Copying the porcupine and rhino models...');
 
@@ -8,36 +9,37 @@ const fixturesDirectory = join(__dirname, '..', 'cypress', 'fixtures');
 
 const engines = [
   {
-    "name": "porcupine",
-    "dir": "keyword_files"
+    name: 'porcupine',
+    dir: 'keyword_files',
   },
   {
-    "name": "rhino",
-    "dir": "contexts"
-  }
+    name: 'rhino',
+    dir: 'contexts',
+  },
 ];
 
-const sourceDirectory = join(
-  __dirname,
-  "..",
-  "..",
-  "..",
-  "resources",
-);
+const sourceDirectory = join(__dirname, '..', '..', '..', 'resources');
 
-const testDataSource = join(
-  sourceDirectory,
-  '.test',
-  'test_data.json'
-);
+const testDataSource = join(sourceDirectory, '.test', 'test_data.json');
 
 try {
   fs.mkdirSync(outputDirectory, { recursive: true });
   fs.copyFileSync(testDataSource, join(outputDirectory, 'test_data.json'));
 
   fs.mkdirSync(join(fixturesDirectory, 'audio_samples'), { recursive: true });
   fs.readdirSync(join(sourceDirectory, 'audio_samples')).forEach(file => {
-    fs.copyFileSync(join(sourceDirectory, 'audio_samples', file), join(fixturesDirectory, 'audio_samples', file));
+    let src = join(sourceDirectory, 'audio_samples', file);
+
+    // Bug in Cypress means we can't read utf-8 file names, so we have to hash them
+    let encodedAudioName = createHash('md5')
+      .update(file.replace('.wav', ''))
+      .digest('hex');
+    let dst = join(
+      fixturesDirectory,
+      'audio_samples',
+      `${encodedAudioName}.wav`
+    );
+    fs.copyFileSync(src, dst);
   });
 
   for (const engine of engines) {
@@ -52,21 +54,41 @@ try {
       sourceDirectory,
       engine.name,
       'resources'
-    )
+    );
 
     fs.mkdirSync(join(outputDirectory, engine.name), { recursive: true });
     fs.readdirSync(paramsSourceDirectory).forEach(file => {
-      fs.copyFileSync(join(paramsSourceDirectory, file), join(outputDirectory, engine.name, file));
+      fs.copyFileSync(
+        join(paramsSourceDirectory, file),
+        join(outputDirectory, engine.name, file)
+      );
     });
 
     fs.mkdirSync(join(outputDirectory, engine.dir), { recursive: true });
     fs.readdirSync(engineSourceDirectory).forEach(folder => {
       if (folder.includes(engine.dir)) {
-        fs.readdirSync(join(engineSourceDirectory, folder, 'wasm')).forEach(file => {
-          fs.copyFileSync(
-            join(engineSourceDirectory, folder, 'wasm', file),
-            join(outputDirectory, engine.dir, file.replace("ā", "a").replace("ō", "o")));
-        });
+        fs.readdirSync(join(engineSourceDirectory, folder, 'wasm')).forEach(
+          file => {
+            let src = join(engineSourceDirectory, folder, 'wasm', file);
+            if (folder === engine.dir) {
+              let dst = join(outputDirectory, engine.dir, file);
+              fs.copyFileSync(src, dst);
+            } else {
+              let fileParts = file.split('_wasm');
+
+              // Bug in Cypress means we can't read utf-8 file names, so we have to hash them
+              let encodedFileName = createHash('md5')
+                .update(fileParts[0])
+                .digest('hex');
+              let dst = join(
+                outputDirectory,
+                engine.dir,
+                `${encodedFileName}_wasm${fileParts[1]}`
+              );
+              fs.copyFileSync(src, dst);
+            }
+          }
+        );
       }
     });
   }