@brickgale/caption-sync 1.0.0

package/dist/utils/caption-generator.js

@@ -0,0 +1,400 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.getAudioDuration = getAudioDuration;
+exports.splitIntoPhrases = splitIntoPhrases;
+exports.countWords = countWords;
+exports.splitByCharacterLimit = splitByCharacterLimit;
+exports.generateCaptionsFromWordTimestamps = generateCaptionsFromWordTimestamps;
+exports.generateCaptions = generateCaptions;
+exports.generateCaptionsFromText = generateCaptionsFromText;
+const fs = __importStar(require("fs"));
+const child_process_1 = require("child_process");
+const config_1 = require("../config");
+const transcription_1 = require("./transcription");
+const providers_1 = require("../providers");
+/**
+ * Get audio duration using ffmpeg
+ *
+ * @param audioPath - Path to audio file
+ * @returns Duration in seconds
+ *
+ * @example
+ * ```typescript
+ * const duration = getAudioDuration('./audio.mp3');
+ * console.log(`Audio is ${duration} seconds long`);
+ * ```
+ */
+function getAudioDuration(audioPath) {
+    try {
+        const output = (0, child_process_1.execSync)(`ffmpeg -i "${audioPath}" 2>&1 | grep "Duration" | cut -d ' ' -f 4 | sed s/,//`, { encoding: 'utf8' });
+        // Parse duration format HH:MM:SS.ms
+        const [hours, minutes, seconds] = output.trim().split(':');
+        const totalSeconds = parseInt(hours) * 3600 + parseInt(minutes) * 60 + parseFloat(seconds);
+        return totalSeconds;
+    }
+    catch (error) {
+        throw new Error(`Failed to get audio duration: ${error}`);
+    }
+}
+/**
+ * Split text into phrases based on punctuation and natural breaks
+ *
+ * @param text - Text to split into phrases
+ * @returns Array of phrases
+ *
+ * @example
+ * ```typescript
+ * const phrases = splitIntoPhrases("Hello world. This is a test.");
+ * // Returns: ["Hello world", "This is a test"]
+ * ```
+ */
+function splitIntoPhrases(text) {
+    const phrases = text
+        .split(/([.!?;]\s+|\n+)/)
+        .filter(phrase => phrase.trim().length > 0)
+        .map(phrase => phrase.trim())
+        .filter(phrase => phrase !== '.' && phrase !== '!' && phrase !== '?' && phrase !== ';');
+    return phrases;
+}
+/**
+ * Count words in a phrase
+ *
+ * @param phrase - Phrase to count words in
+ * @returns Number of words
+ *
+ * @example
+ * ```typescript
+ * const count = countWords("Hello world");
+ * // Returns: 2
+ * ```
+ */
+function countWords(phrase) {
+    return phrase.split(/\s+/).filter(word => word.length > 0).length;
+}
+/**
+ * Split long phrases into smaller chunks based on character limit
+ *
+ * @param phrase - Phrase to split
+ * @param maxCharacters - Maximum characters per chunk
+ * @returns Array of smaller phrases
+ *
+ * @example
+ * ```typescript
+ * const chunks = splitByCharacterLimit("This is a very long sentence that needs to be split", 20);
+ * // Returns: ["This is a very long", "sentence that needs", "to be split"]
+ * ```
+ */
+function splitByCharacterLimit(phrase, maxCharacters) {
+    if (phrase.length <= maxCharacters) {
+        return [phrase];
+    }
+    const words = phrase.split(/\s+/);
+    const chunks = [];
+    let currentChunk = '';
+    for (const word of words) {
+        const testChunk = currentChunk ? `${currentChunk} ${word}` : word;
+        if (testChunk.length <= maxCharacters) {
+            currentChunk = testChunk;
+        }
+        else {
+            if (currentChunk) {
+                chunks.push(currentChunk);
+            }
+            currentChunk = word;
+        }
+    }
+    if (currentChunk) {
+        chunks.push(currentChunk);
+    }
+    return chunks;
+}
+/**
+ * Check if a word is likely a proper noun (name, place, etc.)
+ * Uses simple heuristics: capitalized mid-sentence, common name patterns
+ */
+function isLikelyProperNoun(word, index) {
+    // Skip first word (always capitalized)
+    if (index === 0)
+        return false;
+    // Check if word starts with capital letter
+    const firstChar = word.charAt(0);
+    return firstChar === firstChar.toUpperCase() && /[A-Z]/.test(firstChar);
+}
+/**
+ * Check if a word is a semantic break point (conjunctions, transition words)
+ */
+function isSemanticBreak(word) {
+    const breakWords = new Set([
+        'and',
+        'but',
+        'or',
+        'so',
+        'then',
+        'now',
+        'because',
+        'however',
+        'therefore',
+        'meanwhile',
+        'finally',
+        'also',
+        'plus',
+        'yet',
+    ]);
+    return breakWords.has(word.toLowerCase());
+}
+/**
+ * Calculate pause duration between two consecutive words
+ */
+function getPauseBetween(prevWord, nextWord) {
+    return nextWord.start - prevWord.end;
+}
+/**
+ * Generate captions from word-level timestamps with intelligent splitting
+ *
+ * Groups words into captions based on:
+ * - Character limits (hard constraint)
+ * - Natural pauses in speech (preferred break points)
+ * - Semantic boundaries (proper nouns, conjunctions)
+ * - Minimum words per caption (avoid too-short captions)
+ *
+ * @param words - Array of words with timestamps
+ * @param maxCharacters - Maximum characters per caption
+ * @returns Array of captions with accurate timing
+ *
+ * @example
+ * ```typescript
+ * const words = [
+ *   { word: "Hello", start: 0.0, end: 0.5 },
+ *   { word: "world", start: 0.6, end: 1.0 }
+ * ];
+ * const captions = generateCaptionsFromWordTimestamps(words, 50);
+ * ```
+ */
+function generateCaptionsFromWordTimestamps(words, maxCharacters) {
+    const captions = [];
+    let currentWords = [];
+    let currentText = '';
+    const minPauseDuration = config_1.CaptionConfig.MIN_PAUSE_DURATION;
+    const minWordsPerCaption = config_1.CaptionConfig.MIN_WORDS_PER_CAPTION;
+    for (let i = 0; i < words.length; i++) {
+        const wordObj = words[i];
+        const word = wordObj.word.trim();
+        const testText = currentText ? `${currentText} ${word}` : word;
+        // Calculate pause before this word
+        const pauseBefore = currentWords.length > 0
+            ? getPauseBetween(currentWords[currentWords.length - 1], wordObj)
+            : 0;
+        // Check various break conditions
+        const hasLongPause = pauseBefore >= minPauseDuration;
+        const isProperNoun = isLikelyProperNoun(word, i);
+        const isSemanticBoundary = isSemanticBreak(word);
+        const hasMinWords = currentWords.length >= minWordsPerCaption;
+        const wouldExceedLimit = testText.length > maxCharacters;
+        // Decide whether to break here
+        const shouldBreak = wouldExceedLimit ||
+            (hasMinWords && hasLongPause) ||
+            (hasMinWords &&
+                isProperNoun &&
+                currentText.length > maxCharacters * 0.5) ||
+            (hasMinWords &&
+                isSemanticBoundary &&
+                currentText.length > maxCharacters * 0.6);
+        if (shouldBreak && currentWords.length > 0) {
+            // Save current caption with word timings for highlighting
+            captions.push({
+                text: currentText,
+                start: currentWords[0].start,
+                end: currentWords[currentWords.length - 1].end,
+                words: [...currentWords],
+            });
+            // Start new caption with current word
+            currentText = word;
+            currentWords = [wordObj];
+        }
+        else {
+            // Add word to current caption
+            currentText = testText;
+            currentWords.push(wordObj);
+        }
+    }
+    // Add the last caption with word timings
+    if (currentWords.length > 0) {
+        captions.push({
+            text: currentText,
+            start: currentWords[0].start,
+            end: currentWords[currentWords.length - 1].end,
+            words: [...currentWords],
+        });
+    }
+    return captions;
+}
+/**
+ * Generate caption timing based on weighted distribution
+ *
+ * This function reads a script file and audio file, then generates
+ * caption timing by distributing the audio duration proportionally
+ * based on the word count in each phrase.
+ *
+ * @param options - Configuration options
+ * @param options.scriptPath - Path to script text file
+ * @param options.audioPath - Path to audio file
+ * @param options.maxCharacters - Maximum characters per caption (default: from config)
+ * @param options.useTranscription - Use speech-to-text to verify/correct script (default: false)
+ * @param options.openaiApiKey - OpenAI API key for transcription
+ * @returns Array of captions with timing
+ *
+ * @example
+ * ```typescript
+ * import { generateCaptions } from 'caption-sync';
+ *
+ * const captions = generateCaptions({
+ *   scriptPath: './script.txt',
+ *   audioPath: './audio.mp3',
+ *   maxCharacters: 50
+ * });
+ *
+ * console.log(captions);
+ * // [{ start: 0, end: 2.5, text: "Hello world." }, ...]
+ * ```
+ */
+async function generateCaptions(options) {
+    const { scriptPath, audioPath, maxCharacters = config_1.CaptionConfig.MAX_CHARACTERS_PER_CAPTION, useTranscription = false, openaiApiKey, transcriptionProvider = 'elevenlabs', elevenLabsApiKey, } = options;
+    // Use transcription with word-level timestamps for accurate timing
+    if (useTranscription) {
+        // Determine which API key to use
+        const apiKey = transcriptionProvider === 'elevenlabs' ? elevenLabsApiKey : openaiApiKey;
+        if (!apiKey) {
+            throw new Error(`API key required for ${transcriptionProvider} transcription. ` +
+                `Set ${transcriptionProvider === 'elevenlabs' ? 'ELEVENLABS_API_KEY' : 'OPENAI_API_KEY'} ` +
+                `environment variable or pass the appropriate apiKey option.`);
+        }
+        console.log(`🎤 Transcribing audio with word-level timestamps using ${transcriptionProvider}...`);
+        // Create provider and get word timestamps
+        const provider = (0, providers_1.createTranscriptionProvider)({
+            provider: transcriptionProvider,
+            apiKey,
+        });
+        const words = await provider.transcribeWithTimestamps(audioPath);
+        console.log(`✅ Got ${words.length} words with timestamps`);
+        // Generate captions from word timestamps (accurate timing)
+        const captions = generateCaptionsFromWordTimestamps(words, maxCharacters);
+        console.log(`📝 Generated ${captions.length} captions with accurate timing`);
+        return captions;
+    }
+    // Fallback: Use script file with weighted word-count distribution
+    let scriptText = fs.readFileSync(scriptPath, 'utf-8');
+    scriptText = (0, transcription_1.sanitizeText)(scriptText);
+    // Get audio duration
+    const totalDuration = getAudioDuration(audioPath);
+    // Split into phrases
+    let phrases = splitIntoPhrases(scriptText);
+    if (phrases.length === 0) {
+        throw new Error('No phrases found in script');
+    }
+    // Apply character limit by splitting long phrases
+    const splitPhrases = [];
+    for (const phrase of phrases) {
+        const chunks = splitByCharacterLimit(phrase, maxCharacters);
+        splitPhrases.push(...chunks);
+    }
+    phrases = splitPhrases;
+    // Calculate word counts for weighted timing
+    const wordCounts = phrases.map(countWords);
+    const totalWords = wordCounts.reduce((sum, count) => sum + count, 0);
+    // Generate captions with weighted timing
+    const captions = [];
+    let currentTime = 0;
+    for (let i = 0; i < phrases.length; i++) {
+        const phrase = phrases[i];
+        const wordCount = wordCounts[i];
+        // Calculate duration based on word count proportion
+        const duration = (wordCount / totalWords) * totalDuration;
+        captions.push({
+            start: currentTime,
+            end: currentTime + duration,
+            text: phrase,
+        });
+        currentTime += duration;
+    }
+    return captions;
+}
+/**
+ * Generate captions from text content directly (no file I/O)
+ *
+ * Useful when you already have the script content in memory
+ * and don't want to write it to a file first.
+ *
+ * @param scriptText - The script text content
+ * @param audioPath - Path to audio file
+ * @returns Array of captions with timing
+ *
+ * @example
+ * ```typescript
+ * import { generateCaptionsFromText } from 'caption-sync';
+ *
+ * const script = "Hello world. This is a test.";
+ * const captions = generateCaptionsFromText(script, "./audio.mp3");
+ * ```
+ */
+function generateCaptionsFromText(scriptText, audioPath) {
+    // Get audio duration
+    const totalDuration = getAudioDuration(audioPath);
+    // Split into phrases
+    const phrases = splitIntoPhrases(scriptText);
+    if (phrases.length === 0) {
+        throw new Error('No phrases found in script');
+    }
+    // Calculate word counts for weighted timing
+    const wordCounts = phrases.map(countWords);
+    const totalWords = wordCounts.reduce((sum, count) => sum + count, 0);
+    // Generate captions with weighted timing
+    const captions = [];
+    let currentTime = 0;
+    for (let i = 0; i < phrases.length; i++) {
+        const phrase = phrases[i];
+        const wordCount = wordCounts[i];
+        // Calculate duration based on word count proportion
+        const duration = (wordCount / totalWords) * totalDuration;
+        captions.push({
+            start: currentTime,
+            end: currentTime + duration,
+            text: phrase,
+        });
+        currentTime += duration;
+    }
+    return captions;
+}
+//# sourceMappingURL=caption-generator.js.map

package/dist/utils/caption-generator.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"caption-generator.js","sourceRoot":"","sources":["../../src/lib/utils/caption-generator.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAoBA,4CAgBC;AAcD,4CAWC;AAcD,gCAEC;AAeD,sDA8BC;AAsED,gFAsEC;AA+BD,4CAkGC;AAoBD,4DAuCC;AAlcD,uCAAyB;AACzB,iDAAyC;AAEzC,sCAA0C;AAC1C,mDAA+C;AAC/C,4CAA2D;AAG3D;;;;;;;;;;;GAWG;AACH,SAAgB,gBAAgB,CAAC,SAAiB;IAChD,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,IAAA,wBAAQ,EACrB,cAAc,SAAS,wDAAwD,EAC/E,EAAE,QAAQ,EAAE,MAAM,EAAE,CACrB,CAAC;QAEF,oCAAoC;QACpC,MAAM,CAAC,KAAK,EAAE,OAAO,EAAE,OAAO,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QAC3D,MAAM,YAAY,GAChB,QAAQ,CAAC,KAAK,CAAC,GAAG,IAAI,GAAG,QAAQ,CAAC,OAAO,CAAC,GAAG,EAAE,GAAG,UAAU,CAAC,OAAO,CAAC,CAAC;QAExE,OAAO,YAAY,CAAC;IACtB,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CAAC,iCAAiC,KAAK,EAAE,CAAC,CAAC;IAC5D,CAAC;AACH,CAAC;AAED;;;;;;;;;;;GAWG;AACH,SAAgB,gBAAgB,CAAC,IAAY;IAC3C,MAAM,OAAO,GAAG,IAAI;SACjB,KAAK,CAAC,iBAAiB,CAAC;SACxB,MAAM,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC;SAC1C,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC;SAC5B,MAAM,CACL,MAAM,CAAC,EAAE,CACP,MAAM,KAAK,GAAG,IAAI,MAAM,KAAK,GAAG,IAAI,MAAM,KAAK,GAAG,IAAI,MAAM,KAAK,GAAG,CACvE,CAAC;IAEJ,OAAO,OAAO,CAAC;AACjB,CAAC;AAED;;;;;;;;;;;GAWG;AACH,SAAgB,UAAU,CAAC,MAAc;IACvC,OAAO,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,MAAM,CAAC;AACpE,CAAC;AAED;;;;;;;;;;;;GAYG;AACH,SAAgB,qBAAqB,CACnC,MAAc,EACd,aAAqB;IAErB,IAAI,MAAM,CAAC,MAAM,IAAI,aAAa,EAAE,CAAC;QACnC,OAAO,CAAC,MAAM,CAAC,CAAC;IAClB,CAAC;IAED,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;IAClC,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,IAAI,YAAY,GAAG,EAAE,CAAC;IAEtB,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,SAAS,GAAG,YAAY,CAAC,CAAC,CAAC,GAAG,YAAY,IAAI,IAAI,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;QAElE,IAAI,SAAS,CAAC,MAAM,IAAI,aAAa,EAAE,CAAC;YACtC,YAAY,GAAG,SAAS,CAAC;QAC3B,CAAC;aAAM,CAAC;YACN,IAAI,YAAY,EAAE,CAAC;gBACjB,MAAM,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;YAC5B,CAAC;YACD,YAAY,GAAG,IAAI,CAAC;QACtB,CAAC;IACH,CAAC;IAED,IAAI,YAAY,EAAE,CAAC;QACjB,MAAM,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;IAC5B,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAED;;;GAGG;AACH,SAAS,kBAAkB,CAAC,IAAY,EAAE,KAAa;IACrD,uCAAuC;IACvC,IAAI,KAAK,KAAK,CAAC;QAAE,OAAO,KAAK,CAAC;IAE9B,2CAA2C;IAC3C,MAAM,SAAS,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;IACjC,OAAO,SAAS,KAAK,SAAS,CAAC,WAAW,EAAE,IAAI,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;AAC1E,CAAC;AAED;;GAEG;AACH,SAAS,eAAe,CAAC,IAAY;IACnC,MAAM,UAAU,GAAG,IAAI,GAAG,CAAC;QACzB,KAAK;QACL,KAAK;QACL,IAAI;QACJ,IAAI;QACJ,MAAM;QACN,KAAK;QACL,SAAS;QACT,SAAS;QACT,WAAW;QACX,WAAW;QACX,SAAS;QACT,MAAM;QACN,MAAM;QACN,KAAK;KACN,CAAC,CAAC;IACH,OAAO,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC,CAAC;AAC5C,CAAC;AAED;;GAEG;AACH,SAAS,eAAe,CACtB,QAAuB,EACvB,QAAuB;IAEvB,OAAO,QAAQ,CAAC,KAAK,GAAG,QAAQ,CAAC,GAAG,CAAC;AACvC,CAAC;AAED;;;;;;;;;;;;;;;;;;;;;GAqBG;AACH,SAAgB,kCAAkC,CAChD,KAAsB,EACtB,aAAqB;IAErB,MAAM,QAAQ,GAAc,EAAE,CAAC;IAC/B,IAAI,YAAY,GAAoB,EAAE,CAAC;IACvC,IAAI,WAAW,GAAG,EAAE,CAAC;IAErB,MAAM,gBAAgB,GAAG,sBAAa,CAAC,kBAAkB,CAAC;IAC1D,MAAM,kBAAkB,GAAG,sBAAa,CAAC,qBAAqB,CAAC;IAE/D,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,OAAO,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACzB,MAAM,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC;QACjC,MAAM,QAAQ,GAAG,WAAW,CAAC,CAAC,CAAC,GAAG,WAAW,IAAI,IAAI,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;QAE/D,mCAAmC;QACnC,MAAM,WAAW,GACf,YAAY,CAAC,MAAM,GAAG,CAAC;YACrB,CAAC,CAAC,eAAe,CAAC,YAAY,CAAC,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC,EAAE,OAAO,CAAC;YACjE,CAAC,CAAC,CAAC,CAAC;QAER,iCAAiC;QACjC,MAAM,YAAY,GAAG,WAAW,IAAI,gBAAgB,CAAC;QACrD,MAAM,YAAY,GAAG,kBAAkB,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC;QACjD,MAAM,kBAAkB,GAAG,eAAe,CAAC,IAAI,CAAC,CAAC;QACjD,MAAM,WAAW,GAAG,YAAY,CAAC,MAAM,IAAI,kBAAkB,CAAC;QAC9D,MAAM,gBAAgB,GAAG,QAAQ,CAAC,MAAM,GAAG,aAAa,CAAC;QAEzD,+BAA+B;QAC/B,MAAM,WAAW,GACf,gBAAgB;YAChB,CAAC,WAAW,IAAI,YAAY,CAAC;YAC7B,CAAC,WAAW;gBACV,YAAY;gBACZ,WAAW,CAAC,MAAM,GAAG,aAAa,GAAG,GAAG,CAAC;YAC3C,CAAC,WAAW;gBACV,kBAAkB;gBAClB,WAAW,CAAC,MAAM,GAAG,aAAa,GAAG,GAAG,CAAC,CAAC;QAE9C,IAAI,WAAW,IAAI,YAAY,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC3C,0DAA0D;YAC1D,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI,EAAE,WAAW;gBACjB,KAAK,EAAE,YAAY,CAAC,CAAC,CAAC,CAAC,KAAK;gBAC5B,GAAG,EAAE,YAAY,CAAC,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,GAAG;gBAC9C,KAAK,EAAE,CAAC,GAAG,YAAY,CAAC;aACzB,CAAC,CAAC;YAEH,sCAAsC;YACtC,WAAW,GAAG,IAAI,CAAC;YACnB,YAAY,GAAG,CAAC,OAAO,CAAC,CAAC;QAC3B,CAAC;aAAM,CAAC;YACN,8BAA8B;YAC9B,WAAW,GAAG,QAAQ,CAAC;YACvB,YAAY,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QAC7B,CAAC;IACH,CAAC;IAED,yCAAyC;IACzC,IAAI,YAAY,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC5B,QAAQ,CAAC,IAAI,CAAC;YACZ,IAAI,EAAE,WAAW;YACjB,KAAK,EAAE,YAAY,CAAC,CAAC,CAAC,CAAC,KAAK;YAC5B,GAAG,EAAE,YAAY,CAAC,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,GAAG;YAC9C,KAAK,EAAE,CAAC,GAAG,YAAY,CAAC;SACzB,CAAC,CAAC;IACL,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA4BG;AACI,KAAK,UAAU,gBAAgB,CACpC,OAAgC;IAEhC,MAAM,EACJ,UAAU,EACV,SAAS,EACT,aAAa,GAAG,sBAAa,CAAC,0BAA0B,EACxD,gBAAgB,GAAG,KAAK,EACxB,YAAY,EACZ,qBAAqB,GAAG,YAAY,EACpC,gBAAgB,GACjB,GAAG,OAAO,CAAC;IAEZ,mEAAmE;IACnE,IAAI,gBAAgB,EAAE,CAAC;QACrB,iCAAiC;QACjC,MAAM,MAAM,GACV,qBAAqB,KAAK,YAAY,CAAC,CAAC,CAAC,gBAAgB,CAAC,CAAC,CAAC,YAAY,CAAC;QAE3E,IAAI,CAAC,MAAM,EAAE,CAAC;YACZ,MAAM,IAAI,KAAK,CACb,wBAAwB,qBAAqB,kBAAkB;gBAC7D,OAAO,qBAAqB,KAAK,YAAY,CAAC,CAAC,CAAC,oBAAoB,CAAC,CAAC,CAAC,gBAAgB,GAAG;gBAC1F,6DAA6D,CAChE,CAAC;QACJ,CAAC;QAED,OAAO,CAAC,GAAG,CACT,0DAA0D,qBAAqB,KAAK,CACrF,CAAC;QAEF,0CAA0C;QAC1C,MAAM,QAAQ,GAAG,IAAA,uCAA2B,EAAC;YAC3C,QAAQ,EAAE,qBAAqB;YAC/B,MAAM;SACP,CAAC,CAAC;QAEH,MAAM,KAAK,GAAG,MAAM,QAAQ,CAAC,wBAAwB,CAAC,SAAS,CAAC,CAAC;QAEjE,OAAO,CAAC,GAAG,CAAC,SAAS,KAAK,CAAC,MAAM,wBAAwB,CAAC,CAAC;QAE3D,2DAA2D;QAC3D,MAAM,QAAQ,GAAG,kCAAkC,CAAC,KAAK,EAAE,aAAa,CAAC,CAAC;QAE1E,OAAO,CAAC,GAAG,CACT,gBAAgB,QAAQ,CAAC,MAAM,gCAAgC,CAChE,CAAC;QACF,OAAO,QAAQ,CAAC;IAClB,CAAC;IAED,kEAAkE;IAClE,IAAI,UAAU,GAAG,EAAE,CAAC,YAAY,CAAC,UAAU,EAAE,OAAO,CAAC,CAAC;IACtD,UAAU,GAAG,IAAA,4BAAY,EAAC,UAAU,CAAC,CAAC;IAEtC,qBAAqB;IACrB,MAAM,aAAa,GAAG,gBAAgB,CAAC,SAAS,CAAC,CAAC;IAElD,qBAAqB;IACrB,IAAI,OAAO,GAAG,gBAAgB,CAAC,UAAU,CAAC,CAAC;IAE3C,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACzB,MAAM,IAAI,KAAK,CAAC,4BAA4B,CAAC,CAAC;IAChD,CAAC;IAED,kDAAkD;IAClD,MAAM,YAAY,GAAa,EAAE,CAAC;IAClC,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;QAC7B,MAAM,MAAM,GAAG,qBAAqB,CAAC,MAAM,EAAE,aAAa,CAAC,CAAC;QAC5D,YAAY,CAAC,IAAI,CAAC,GAAG,MAAM,CAAC,CAAC;IAC/B,CAAC;IAED,OAAO,GAAG,YAAY,CAAC;IAEvB,4CAA4C;IAC5C,MAAM,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC;IAC3C,MAAM,UAAU,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,KAAK,EAAE,EAAE,CAAC,GAAG,GAAG,KAAK,EAAE,CAAC,CAAC,CAAC;IAErE,yCAAyC;IACzC,MAAM,QAAQ,GAAc,EAAE,CAAC;IAC/B,IAAI,WAAW,GAAG,CAAC,CAAC;IAEpB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACxC,MAAM,MAAM,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC;QAC1B,MAAM,SAAS,GAAG,UAAU,CAAC,CAAC,CAAC,CAAC;QAEhC,oDAAoD;QACpD,MAAM,QAAQ,GAAG,CAAC,SAAS,GAAG,UAAU,CAAC,GAAG,aAAa,CAAC;QAE1D,QAAQ,CAAC,IAAI,CAAC;YACZ,KAAK,EAAE,WAAW;YAClB,GAAG,EAAE,WAAW,GAAG,QAAQ;YAC3B,IAAI,EAAE,MAAM;SACb,CAAC,CAAC;QAEH,WAAW,IAAI,QAAQ,CAAC;IAC1B,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED;;;;;;;;;;;;;;;;;GAiBG;AACH,SAAgB,wBAAwB,CACtC,UAAkB,EAClB,SAAiB;IAEjB,qBAAqB;IACrB,MAAM,aAAa,GAAG,gBAAgB,CAAC,SAAS,CAAC,CAAC;IAElD,qBAAqB;IACrB,MAAM,OAAO,GAAG,gBAAgB,CAAC,UAAU,CAAC,CAAC;IAE7C,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACzB,MAAM,IAAI,KAAK,CAAC,4BAA4B,CAAC,CAAC;IAChD,CAAC;IAED,4CAA4C;IAC5C,MAAM,UAAU,GAAG,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC;IAC3C,MAAM,UAAU,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,KAAK,EAAE,EAAE,CAAC,GAAG,GAAG,KAAK,EAAE,CAAC,CAAC,CAAC;IAErE,yCAAyC;IACzC,MAAM,QAAQ,GAAc,EAAE,CAAC;IAC/B,IAAI,WAAW,GAAG,CAAC,CAAC;IAEpB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACxC,MAAM,MAAM,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC;QAC1B,MAAM,SAAS,GAAG,UAAU,CAAC,CAAC,CAAC,CAAC;QAEhC,oDAAoD;QACpD,MAAM,QAAQ,GAAG,CAAC,SAAS,GAAG,UAAU,CAAC,GAAG,aAAa,CAAC;QAE1D,QAAQ,CAAC,IAAI,CAAC;YACZ,KAAK,EAAE,WAAW;YAClB,GAAG,EAAE,WAAW,GAAG,QAAQ;YAC3B,IAAI,EAAE,MAAM;SACb,CAAC,CAAC;QAEH,WAAW,IAAI,QAAQ,CAAC;IAC1B,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/utils/transcription.d.ts

@@ -0,0 +1,96 @@
+import type { TranscribeAudioOptions, TextComparisonResult } from '../types';
+/**
+ * Word with timestamp information
+ */
+export interface WordTimestamp {
+    word: string;
+    start: number;
+    end: number;
+}
+/**
+ * Transcribe audio to text using OpenAI Whisper
+ *
+ * @param options - Transcription options
+ * @returns Transcribed text
+ *
+ * @example
+ * ```typescript
+ * const text = await transcribeAudio({
+ *   audioPath: './audio.mp3',
+ *   openaiApiKey: process.env.OPENAI_API_KEY!,
+ * });
+ * console.log('Transcribed:', text);
+ * ```
+ */
+export declare function transcribeAudio(options: TranscribeAudioOptions): Promise<string>;
+/**
+ * Transcribe audio with word-level timestamps using OpenAI Whisper
+ *
+ * @param options - Transcription options
+ * @returns Array of words with timestamps
+ *
+ * @example
+ * ```typescript
+ * const words = await transcribeAudioWithTimestamps({
+ *   audioPath: './audio.mp3',
+ *   openaiApiKey: process.env.OPENAI_API_KEY!,
+ * });
+ * console.log('First word:', words[0]);
+ * // { word: "Hello", start: 0.5, end: 0.8 }
+ * ```
+ */
+export declare function transcribeAudioWithTimestamps(options: TranscribeAudioOptions): Promise<WordTimestamp[]>;
+/**
+ * Sanitize text by removing unwanted characters
+ * Keeps only alphanumeric characters, spaces, and basic punctuation
+ *
+ * @param text - Text to sanitize
+ * @returns Sanitized text
+ *
+ * @example
+ * ```typescript
+ * const clean = sanitizeText("Hello—world! [test]");
+ * // Returns: "Hello world! test"
+ * ```
+ */
+export declare function sanitizeText(text: string): string;
+/**
+ * Compare original script with transcribed audio
+ *
+ * @param originalText - Original script text
+ * @param transcribedText - Transcribed audio text
+ * @returns Comparison result with differences and similarity score
+ *
+ * @example
+ * ```typescript
+ * const result = compareTexts(originalScript, transcribedText);
+ * console.log(`Similarity: ${(result.similarity * 100).toFixed(1)}%`);
+ * console.log('Differences:', result.differences);
+ * ```
+ */
+export declare function compareTexts(originalText: string, transcribedText: string): TextComparisonResult;
+/**
+ * Generate corrected script based on transcription
+ *
+ * @param scriptPath - Path to original script file
+ * @param audioPath - Path to audio file
+ * @param openaiApiKey - OpenAI API key for transcription
+ * @returns Comparison result with corrections
+ *
+ * @example
+ * ```typescript
+ * const result = await generateCorrections({
+ *   scriptPath: './script.txt',
+ *   audioPath: './audio.mp3',
+ *   openaiApiKey: process.env.OPENAI_API_KEY!
+ * });
+ *
+ * console.log('Corrections needed:', result.differences);
+ * ```
+ */
+export declare function generateCorrections(options: {
+    scriptPath: string;
+    audioPath: string;
+    openaiApiKey: string;
+}): Promise<TextComparisonResult>;
+//# sourceMappingURL=transcription.d.ts.map

package/dist/utils/transcription.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"transcription.d.ts","sourceRoot":"","sources":["../../src/lib/utils/transcription.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,sBAAsB,EAAE,oBAAoB,EAAE,MAAM,UAAU,CAAC;AAG7E;;GAEG;AACH,MAAM,WAAW,aAAa;IAC5B,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,MAAM,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;CACb;AAED;;;;;;;;;;;;;;GAcG;AACH,wBAAsB,eAAe,CACnC,OAAO,EAAE,sBAAsB,GAC9B,OAAO,CAAC,MAAM,CAAC,CAkCjB;AAED;;;;;;;;;;;;;;;GAeG;AACH,wBAAsB,6BAA6B,CACjD,OAAO,EAAE,sBAAsB,GAC9B,OAAO,CAAC,aAAa,EAAE,CAAC,CA8C1B;AAED;;;;;;;;;;;;GAYG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAUjD;AAgED;;;;;;;;;;;;;GAaG;AACH,wBAAgB,YAAY,CAC1B,YAAY,EAAE,MAAM,EACpB,eAAe,EAAE,MAAM,GACtB,oBAAoB,CAsBtB;AAED;;;;;;;;;;;;;;;;;;GAkBG;AACH,wBAAsB,mBAAmB,CAAC,OAAO,EAAE;IACjD,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,YAAY,EAAE,MAAM,CAAC;CACtB,GAAG,OAAO,CAAC,oBAAoB,CAAC,CAgChC"}