@remotion/openai-whisper 4.0.243 → 4.0.244

package/dist/openai-whisper-api-to-captions.js

@@ -8,7 +8,7 @@ const openAiWhisperApiToCaptions = ({ transcription, }) => {
     }
     let remainingText = transcription.text;
     for (const word of transcription.words) {
-        const match = new RegExp(`^([\\s\\.]{0,4})${word.word}([\\?,\\.\\%\\–\\!\\;\\:\\'\\"\\-\\_\\(\\)\\[\\]\\{\\}\\@\\#\\$\\^\\&\\*\\+\\=\\/\\|\\<\\>\\~\`]{0,3})?`).exec(remainingText);
+        const match = new RegExp(`^([\\s\\.]{0,4})${word.word.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}([\\?,\\.\\%\\–\\!\\;\\:\\'\\"\\-\\_\\(\\)\\[\\]\\{\\}\\@\\#\\$\\^\\&\\*\\+\\=\\/\\|\\<\\>\\~\`]{0,3})?`).exec(remainingText);
         if (!match) {
             throw new Error(`Unable to parse punctuation from OpenAI Whisper output. Could not find word "${word.word}" in text "${remainingText.slice(0, 100)}". File an issue under https://remotion.dev/issue to ask for a fix.`);
         }

package/dist/test/get-and-convert.test.js

@@ -88,7 +88,7 @@ if (!process.env.CI) {
     (0, bun_test_1.test)('Should convert to captions', async () => {
         const openai = new openai_1.default();
         const transcription = await openai.audio.transcriptions.create({
-            file: fs_1.default.createReadStream(path_1.default.join(__dirname, '..', '..', '..', 'template-audiogram', 'public', 'audio.mp3')),
+            file: fs_1.default.createReadStream(path_1.default.join(__dirname, '..', '..', '..', 'template-audiogram', 'public', 'audio.wav')),
             model: 'whisper-1',
             response_format: 'verbose_json',
             prompt: 'Hello, welcome to my lecture.',

package/dist/test/special-chars.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/test/special-chars.test.js

@@ -0,0 +1,139 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+const bun_test_1 = require("bun:test");
+const openai_whisper_api_to_captions_1 = require("../openai-whisper-api-to-captions");
+const text = ' $500 that you can spend on a computer right now.';
+const transcript = {
+    task: 'transcribe',
+    text,
+    words: [
+        {
+            start: 1,
+            end: 2,
+            word: '$500',
+        },
+        {
+            start: 3,
+            end: 4,
+            word: 'that',
+        },
+        {
+            start: 5,
+            end: 6,
+            word: 'you',
+        },
+        {
+            start: 7,
+            end: 8,
+            word: 'can',
+        },
+        {
+            start: 9,
+            end: 10,
+            word: 'spend',
+        },
+        {
+            start: 11,
+            end: 12,
+            word: 'on',
+        },
+        {
+            start: 13,
+            end: 14,
+            word: 'a',
+        },
+        {
+            start: 15,
+            end: 16,
+            word: 'computer',
+        },
+        {
+            start: 17,
+            end: 18,
+            word: 'right',
+        },
+        {
+            start: 19,
+            end: 20,
+            word: 'now.',
+        },
+    ],
+    duration: 4,
+    language: 'english',
+};
+(0, bun_test_1.test)('should not crash', () => {
+    (0, bun_test_1.expect)((0, openai_whisper_api_to_captions_1.openAiWhisperApiToCaptions)({ transcription: transcript })).toEqual({
+        captions: [
+            {
+                confidence: null,
+                endMs: 2000,
+                startMs: 1000,
+                text: ' $500',
+                timestampMs: 1500,
+            },
+            {
+                confidence: null,
+                endMs: 4000,
+                startMs: 3000,
+                text: ' that',
+                timestampMs: 3500,
+            },
+            {
+                confidence: null,
+                endMs: 6000,
+                startMs: 5000,
+                text: ' you',
+                timestampMs: 5500,
+            },
+            {
+                confidence: null,
+                endMs: 8000,
+                startMs: 7000,
+                text: ' can',
+                timestampMs: 7500,
+            },
+            {
+                confidence: null,
+                endMs: 10000,
+                startMs: 9000,
+                text: ' spend',
+                timestampMs: 9500,
+            },
+            {
+                confidence: null,
+                endMs: 12000,
+                startMs: 11000,
+                text: ' on',
+                timestampMs: 11500,
+            },
+            {
+                confidence: null,
+                endMs: 14000,
+                startMs: 13000,
+                text: ' a',
+                timestampMs: 13500,
+            },
+            {
+                confidence: null,
+                endMs: 16000,
+                startMs: 15000,
+                text: ' computer',
+                timestampMs: 15500,
+            },
+            {
+                confidence: null,
+                endMs: 18000,
+                startMs: 17000,
+                text: ' right',
+                timestampMs: 17500,
+            },
+            {
+                confidence: null,
+                endMs: 20000,
+                startMs: 19000,
+                text: ' now.',
+                timestampMs: 19500,
+            },
+        ],
+    });
+});

package/package.json

@@ -3,7 +3,7 @@
     "url": "https://github.com/remotion-dev/remotion/tree/main/packages/openai-whisper"
   },
   "name": "@remotion/openai-whisper",
-  "version": "4.0.243",
+  "version": "4.0.244",
   "description": "Work with the output of the OpenAI Whisper API",
   "main": "dist/index.js",
   "sideEffects": false,
@@ -16,13 +16,13 @@
   "author": "Jonny Burger <jonny@remotion.dev>",
   "license": "MIT",
   "dependencies": {
-    "@remotion/captions": "4.0.243"
+    "@remotion/captions": "4.0.244"
   },
   "peerDependencies": {},
   "devDependencies": {
     "openai": "4.67.1",
     "eslint": "9.14.0",
-    "@remotion/eslint-config-internal": "4.0.243"
+    "@remotion/eslint-config-internal": "4.0.244"
   },
   "keywords": [
     "remotion"