Fix transcription (#1137)

* upgrade deps

* fix generate transcription

* validate transcription before download from online

1000-hours/package.json

@@ -8,7 +8,7 @@
     "markdown-it-sub": "^2.0.0",
     "markdown-it-sup": "^2.0.0",
     "mermaid": "^11.3.0",
-    "sass": "^1.80.1",
+    "sass": "^1.80.3",
     "vitepress": "^1.4.1",
     "vitepress-plugin-mermaid": "^2.0.17",
     "vue": "^3.5.12"

1000h-portal/package.json

@@ -19,7 +19,7 @@
   "devDependencies": {
     "autoprefixer": "^10.4.20",
     "postcss": "^8.4.47",
-    "sass": "^1.80.1",
+    "sass": "^1.80.3",
     "tailwindcss": "^3.4.14"
   }
 }

enjoy/package.json

@@ -49,10 +49,10 @@
     "@types/fluent-ffmpeg": "^2.1.26",
     "@types/html-to-text": "^9.0.4",
     "@types/intl-tel-input": "^18.1.4",
-    "@types/lodash": "^4.17.10",
+    "@types/lodash": "^4.17.12",
     "@types/mark.js": "^8.11.12",
     "@types/mustache": "^4.2.5",
-    "@types/node": "^22.7.6",
+    "@types/node": "^22.7.7",
     "@types/prop-types": "^15.7.13",
     "@types/rails__actioncable": "^6.1.11",
     "@types/react": "^18.3.11",
@@ -61,14 +61,14 @@
     "@types/unzipper": "^0.10.10",
     "@types/validator": "^13.12.2",
     "@types/wavesurfer.js": "^6.0.12",
-    "@typescript-eslint/eslint-plugin": "^8.9.0",
-    "@typescript-eslint/parser": "^8.9.0",
-    "@vitejs/plugin-react": "^4.3.2",
+    "@typescript-eslint/eslint-plugin": "^8.10.0",
+    "@typescript-eslint/parser": "^8.10.0",
+    "@vitejs/plugin-react": "^4.3.3",
     "autoprefixer": "^10.4.20",
     "electron": "^33.0.1",
     "electron-devtools-installer": "^3.2.0",
     "electron-playwright-helpers": "^1.7.1",
-    "eslint": "^9.12.0",
+    "eslint": "^9.13.0",
     "eslint-import-resolver-typescript": "^3.6.3",
     "eslint-plugin-import": "^2.31.0",
     "flora-colossus": "^2.0.0",
@@ -139,7 +139,7 @@
     "dayjs": "^1.11.13",
     "decamelize": "^6.0.0",
     "decamelize-keys": "^2.0.1",
-    "echogarden": "^1.8.5",
+    "echogarden": "^1.8.7",
     "electron-context-menu": "^4.0.4",
     "electron-log": "^5.2.0",
     "electron-settings": "^4.0.4",
@@ -149,7 +149,7 @@
     "fs-extra": "^11.2.0",
     "html-to-text": "^9.0.5",
     "https-proxy-agent": "^7.0.5",
-    "i18next": "^23.16.0",
+    "i18next": "^23.16.1",
     "intl-tel-input": "^24.6.0",
     "js-md5": "^0.8.3",
     "langchain": "^0.3.2",
@@ -160,17 +160,17 @@
     "microsoft-cognitiveservices-speech-sdk": "^1.41.0",
     "mustache": "^4.2.0",
     "next-themes": "^0.3.0",
-    "openai": "^4.67.3",
+    "openai": "^4.68.1",
     "pitchfinder": "^2.3.2",
     "postcss": "^8.4.47",
     "proxy-agent": "^6.4.0",
     "react": "^18.3.1",
-    "react-activity-calendar": "^2.6.2",
+    "react-activity-calendar": "^2.6.3",
     "react-audio-visualize": "^1.2.0",
     "react-audio-voice-recorder": "^2.2.0",
     "react-dom": "^18.3.1",
     "react-frame-component": "^5.2.7",
-    "react-hook-form": "^7.53.0",
+    "react-hook-form": "^7.53.1",
     "react-hotkeys-hook": "^4.5.1",
     "react-i18next": "^15.0.3",
     "react-markdown": "^9.0.1",
@@ -189,7 +189,7 @@
     "umzug": "^3.8.2",
     "unzipper": "^0.12.3",
     "update-electron-app": "^3.0.0",
-    "wavesurfer.js": "^7.8.6",
+    "wavesurfer.js": "^7.8.7",
     "zod": "^3.23.8",
     "zod-to-json-schema": "^3.23.3"
   }

enjoy/src/renderer/hooks/use-transcriptions.tsx

@@ -94,7 +94,13 @@ export const useTranscriptions = (media: AudioType | VideoType) => {
         items: 10,
       });
       if (result.transcriptions.length) {
-        return result.transcriptions[0];
+        for (const tr of result.transcriptions) {
+          if (validateTranscription(tr)) {
+            return tr;
+          } else {
+            console.warn(`Invalid transcription: ${tr.id}`);
+          }
+        }
       } else {
         return null;
       }
@@ -115,7 +121,7 @@ export const useTranscriptions = (media: AudioType | VideoType) => {
       originalText,
       language = learningLanguage,
       service = sttEngine,
-      model,
+      model: whisperModel,
       isolate = false,
     } = params || {};
     setService(service);
@@ -138,6 +144,7 @@ export const useTranscriptions = (media: AudioType | VideoType) => {
         {
           targetId: media.id,
           targetType: media.mediaType,
+          model: whisperModel,
           originalText,
           language,
           service,
@@ -254,6 +261,21 @@ export const useTranscriptions = (media: AudioType | VideoType) => {
     return timeline;
   };
 
+  const validateTranscription = (transcription: TranscriptionType) => {
+    if (!transcription) return;
+
+    const { timeline, transcript } = transcription.result;
+    if (!timeline || !transcript) {
+      return false;
+    }
+
+    if (timeline[0]?.type !== "sentence") {
+      return false;
+    }
+
+    return true;
+  };
+
   /*
    * find or create transcription
    */