@ninebix/nmt-system 1.0.2 → 1.0.3

package/dist/core/language-analyzers.js

@@ -1,365 +0,0 @@
-/**
- * Language-specific Analyzers
- * 언어별 형태소 분석 및 청킹 최적화
- *
- * @module core/language-analyzers
- */
-/**
- * Korean Analyzer
- * 한국어 형태소 분석기
- */
-export class KoreanAnalyzer {
-    language = 'ko';
-    // 한국어 조사 목록
-    particles = new Set([
-        '은', '는', '이', '가', '을', '를', '에', '에서', '에게', '께',
-        '의', '로', '으로', '와', '과', '랑', '이랑', '도', '만', '부터',
-        '까지', '보다', '처럼', '같이', '마다', '밖에', '뿐', '조차'
-    ]);
-    // 한국어 어미 목록
-    suffixes = new Set([
-        '다', '니다', '습니다', '세요', '요', '네', '죠', '지', '고',
-        '면', '어서', '아서', '니까', '는데', '던', '었', '았', '겠'
-    ]);
-    // 문장 종결 패턴
-    sentenceEnders = /[.!?。！？]\s*/g;
-    tokenize(text) {
-        const tokens = [];
-        const words = text.split(/\s+/);
-        for (const word of words) {
-            if (!word)
-                continue;
-            // 조사 분리
-            let remaining = word;
-            let foundParticle = false;
-            for (const particle of this.particles) {
-                if (remaining.endsWith(particle) && remaining.length > particle.length) {
-                    const stem = remaining.slice(0, -particle.length);
-                    tokens.push({ text: stem, type: 'word' });
-                    tokens.push({ text: particle, type: 'particle' });
-                    foundParticle = true;
-                    break;
-                }
-            }
-            if (!foundParticle) {
-                // 어미 분리 시도
-                let foundSuffix = false;
-                for (const suffix of this.suffixes) {
-                    if (remaining.endsWith(suffix) && remaining.length > suffix.length) {
-                        const stem = remaining.slice(0, -suffix.length);
-                        tokens.push({ text: stem, type: 'word' });
-                        tokens.push({ text: suffix, type: 'suffix' });
-                        foundSuffix = true;
-                        break;
-                    }
-                }
-                if (!foundSuffix) {
-                    tokens.push({ text: remaining, type: 'word' });
-                }
-            }
-        }
-        return tokens;
-    }
-    splitSentences(text) {
-        return text.split(this.sentenceEnders).filter(s => s.trim().length > 0);
-    }
-    splitPhrases(text) {
-        // 쉼표, 접속사로 분리
-        return text.split(/[,，、]|\s+(그리고|그러나|하지만|그래서|또한)\s+/g)
-            .filter(p => p && p.trim().length > 0);
-    }
-    splitWords(text) {
-        return text.split(/\s+/).filter(w => w.trim().length > 0);
-    }
-    splitMorphemes(text) {
-        return this.tokenize(text).map(t => t.text);
-    }
-}
-/**
- * English Analyzer
- * 영어 형태소 분석기
- */
-export class EnglishAnalyzer {
-    language = 'en';
-    // Common English suffixes
-    suffixes = ['ing', 'ed', 'ly', 'tion', 'sion', 'ness', 'ment', 'able', 'ible'];
-    // Stop words
-    stopWords = new Set([
-        'the', 'a', 'an', 'is', 'are', 'was', 'were', 'be', 'been', 'being',
-        'have', 'has', 'had', 'do', 'does', 'did', 'will', 'would', 'could',
-        'should', 'may', 'might', 'must', 'shall', 'can', 'of', 'at', 'by',
-        'for', 'with', 'about', 'against', 'between', 'into', 'through',
-        'during', 'before', 'after', 'above', 'below', 'to', 'from', 'up',
-        'down', 'in', 'out', 'on', 'off', 'over', 'under', 'again', 'further',
-        'then', 'once', 'and', 'but', 'or', 'nor', 'so', 'yet'
-    ]);
-    tokenize(text) {
-        const tokens = [];
-        const words = text.toLowerCase().split(/\s+/);
-        for (const word of words) {
-            if (!word)
-                continue;
-            // Remove punctuation
-            const cleanWord = word.replace(/[^\w]/g, '');
-            if (this.stopWords.has(cleanWord)) {
-                tokens.push({ text: cleanWord, type: 'particle' });
-            }
-            else {
-                // Try to find suffix
-                let foundSuffix = false;
-                for (const suffix of this.suffixes) {
-                    if (cleanWord.endsWith(suffix) && cleanWord.length > suffix.length + 2) {
-                        const stem = cleanWord.slice(0, -suffix.length);
-                        tokens.push({ text: stem, type: 'word', lemma: stem });
-                        tokens.push({ text: suffix, type: 'suffix' });
-                        foundSuffix = true;
-                        break;
-                    }
-                }
-                if (!foundSuffix) {
-                    tokens.push({ text: cleanWord, type: 'word' });
-                }
-            }
-        }
-        return tokens;
-    }
-    splitSentences(text) {
-        return text.split(/(?<=[.!?])\s+/).filter(s => s.trim().length > 0);
-    }
-    splitPhrases(text) {
-        return text.split(/[,;:]|\s+(and|but|or|because|although|however)\s+/gi)
-            .filter(p => p && p.trim().length > 0 && p.length > 3);
-    }
-    splitWords(text) {
-        return text.split(/\s+/).filter(w => w.trim().length > 0);
-    }
-    splitMorphemes(text) {
-        return this.tokenize(text).map(t => t.text);
-    }
-}
-/**
- * Japanese Analyzer
- * 일본어 형태소 분석기
- */
-export class JapaneseAnalyzer {
-    language = 'ja';
-    // Unicode ranges
-    hiraganaRange = /[\u3040-\u309F]/;
-    katakanaRange = /[\u30A0-\u30FF]/;
-    kanjiRange = /[\u4E00-\u9FAF]/;
-    tokenize(text) {
-        const tokens = [];
-        let currentToken = '';
-        let currentType = 'word';
-        for (const char of text) {
-            let charType = 'word';
-            if (this.hiraganaRange.test(char)) {
-                charType = 'hiragana';
-            }
-            else if (this.katakanaRange.test(char)) {
-                charType = 'katakana';
-            }
-            else if (this.kanjiRange.test(char)) {
-                charType = 'kanji';
-            }
-            else if (/\s/.test(char)) {
-                charType = 'whitespace';
-            }
-            else if (/[。！？、]/.test(char)) {
-                charType = 'punctuation';
-            }
-            if (charType !== currentType && currentToken) {
-                tokens.push({ text: currentToken, type: currentType });
-                currentToken = '';
-            }
-            if (charType !== 'whitespace') {
-                currentToken += char;
-            }
-            currentType = charType;
-        }
-        if (currentToken) {
-            tokens.push({ text: currentToken, type: currentType });
-        }
-        return tokens;
-    }
-    splitSentences(text) {
-        return text.split(/[。！？]\s*/).filter(s => s.trim().length > 0);
-    }
-    splitPhrases(text) {
-        return text.split(/[、，]\s*/).filter(p => p.trim().length > 0);
-    }
-    splitWords(text) {
-        // Japanese doesn't use spaces, so we split by character type changes
-        const tokens = this.tokenize(text);
-        return tokens.filter(t => t.type !== 'punctuation').map(t => t.text);
-    }
-    splitMorphemes(text) {
-        return this.tokenize(text).map(t => t.text);
-    }
-}
-/**
- * Chinese Analyzer
- * 중국어 형태소 분석기
- */
-export class ChineseAnalyzer {
-    language = 'zh';
-    // Common Chinese particles/function words
-    particles = new Set([
-        '的', '了', '是', '在', '有', '和', '与', '或', '但', '而',
-        '就', '也', '都', '还', '又', '才', '只', '已', '把', '被',
-        '让', '给', '向', '从', '到', '对', '为', '以', '因', '所'
-    ]);
-    tokenize(text) {
-        const tokens = [];
-        for (const char of text) {
-            if (/\s/.test(char))
-                continue;
-            if (/[。！？，、]/.test(char)) {
-                tokens.push({ text: char, type: 'punctuation' });
-            }
-            else if (this.particles.has(char)) {
-                tokens.push({ text: char, type: 'particle' });
-            }
-            else {
-                tokens.push({ text: char, type: 'hanzi' });
-            }
-        }
-        return tokens;
-    }
-    splitSentences(text) {
-        return text.split(/[。！？]\s*/).filter(s => s.trim().length > 0);
-    }
-    splitPhrases(text) {
-        return text.split(/[，、；]\s*/).filter(p => p.trim().length > 0);
-    }
-    splitWords(text) {
-        // Basic bigram approach for Chinese
-        const chars = [...text].filter(c => !/\s/.test(c) && !/[。！？，、；]/.test(c));
-        const words = [];
-        for (let i = 0; i < chars.length; i += 2) {
-            if (i + 1 < chars.length) {
-                words.push(chars[i] + chars[i + 1]);
-            }
-            else {
-                words.push(chars[i]);
-            }
-        }
-        return words;
-    }
-    splitMorphemes(text) {
-        // Each character is a morpheme in Chinese
-        return [...text].filter(c => !/\s/.test(c));
-    }
-}
-/**
- * Auto-detect language and return appropriate analyzer
- */
-export function detectLanguage(text) {
-    // Count character types
-    let korean = 0, japanese = 0, chinese = 0, english = 0, total = 0;
-    for (const char of text) {
-        if (/\s/.test(char))
-            continue;
-        total++;
-        // Korean Hangul
-        if (/[\uAC00-\uD7AF\u1100-\u11FF\u3130-\u318F]/.test(char)) {
-            korean++;
-        }
-        // Japanese (Hiragana, Katakana)
-        else if (/[\u3040-\u309F\u30A0-\u30FF]/.test(char)) {
-            japanese++;
-        }
-        // CJK Unified Ideographs (shared by Chinese/Japanese/Korean)
-        else if (/[\u4E00-\u9FAF]/.test(char)) {
-            chinese++; // Assume Chinese if no kana
-        }
-        // Latin
-        else if (/[a-zA-Z]/.test(char)) {
-            english++;
-        }
-    }
-    if (total === 0)
-        return 'en';
-    const koRatio = korean / total;
-    const jaRatio = japanese / total;
-    const zhRatio = chinese / total;
-    const enRatio = english / total;
-    // Determine language
-    if (koRatio > 0.3)
-        return 'ko';
-    if (jaRatio > 0.1)
-        return 'ja'; // Even small amount of kana indicates Japanese
-    if (zhRatio > 0.5)
-        return 'zh';
-    if (enRatio > 0.5)
-        return 'en';
-    // Default fallback
-    if (koRatio > jaRatio && koRatio > zhRatio)
-        return 'ko';
-    if (jaRatio > zhRatio)
-        return 'ja';
-    if (zhRatio > 0)
-        return 'zh';
-    return 'en';
-}
-/**
- * Get analyzer for language
- */
-export function getAnalyzer(language, text) {
-    let lang = language;
-    if (language === 'auto' && text) {
-        lang = detectLanguage(text);
-    }
-    switch (lang) {
-        case 'ko':
-            return new KoreanAnalyzer();
-        case 'ja':
-            return new JapaneseAnalyzer();
-        case 'zh':
-            return new ChineseAnalyzer();
-        case 'en':
-        default:
-            return new EnglishAnalyzer();
-    }
-}
-/**
- * Multi-language tokenizer
- */
-export class MultiLanguageTokenizer {
-    analyzers = new Map();
-    constructor() {
-        this.analyzers.set('ko', new KoreanAnalyzer());
-        this.analyzers.set('en', new EnglishAnalyzer());
-        this.analyzers.set('ja', new JapaneseAnalyzer());
-        this.analyzers.set('zh', new ChineseAnalyzer());
-    }
-    /**
-     * Tokenize with auto language detection
-     */
-    tokenize(text) {
-        const language = detectLanguage(text);
-        const analyzer = this.analyzers.get(language) || new EnglishAnalyzer();
-        return {
-            language,
-            tokens: analyzer.tokenize(text)
-        };
-    }
-    /**
-     * Split into all levels
-     */
-    analyze(text) {
-        const language = detectLanguage(text);
-        const analyzer = this.analyzers.get(language) || new EnglishAnalyzer();
-        return {
-            language,
-            sentences: analyzer.splitSentences(text),
-            phrases: analyzer.splitPhrases(text),
-            words: analyzer.splitWords(text),
-            morphemes: analyzer.splitMorphemes(text),
-            tokens: analyzer.tokenize(text)
-        };
-    }
-}
-// Export singleton
-export const multiLangTokenizer = new MultiLanguageTokenizer();
-//# sourceMappingURL=language-analyzers.js.map

package/dist/core/language-analyzers.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"language-analyzers.js","sourceRoot":"","sources":["../../src/core/language-analyzers.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AA0CH;;;GAGG;AACH,MAAM,OAAO,cAAc;IACzB,QAAQ,GAAa,IAAI,CAAC;IAE1B,YAAY;IACJ,SAAS,GAAG,IAAI,GAAG,CAAC;QAC1B,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,IAAI,EAAE,GAAG;QAClD,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI;QACnD,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,GAAG,EAAE,IAAI;KAC9C,CAAC,CAAC;IAEH,YAAY;IACJ,QAAQ,GAAG,IAAI,GAAG,CAAC;QACzB,GAAG,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG;QAC/C,GAAG,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG;KAChD,CAAC,CAAC;IAEH,WAAW;IACH,cAAc,GAAG,cAAc,CAAC;IAExC,QAAQ,CAAC,IAAY;QACnB,MAAM,MAAM,GAAY,EAAE,CAAC;QAC3B,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;QAEhC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,IAAI,CAAC,IAAI;gBAAE,SAAS;YAEpB,QAAQ;YACR,IAAI,SAAS,GAAG,IAAI,CAAC;YACrB,IAAI,aAAa,GAAG,KAAK,CAAC;YAE1B,KAAK,MAAM,QAAQ,IAAI,IAAI,CAAC,SAAS,EAAE,CAAC;gBACtC,IAAI,SAAS,CAAC,QAAQ,CAAC,QAAQ,CAAC,IAAI,SAAS,CAAC,MAAM,GAAG,QAAQ,CAAC,MAAM,EAAE,CAAC;oBACvE,MAAM,IAAI,GAAG,SAAS,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;oBAClD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC;oBAC1C,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC,CAAC;oBAClD,aAAa,GAAG,IAAI,CAAC;oBACrB,MAAM;gBACR,CAAC;YACH,CAAC;YAED,IAAI,CAAC,aAAa,EAAE,CAAC;gBACnB,WAAW;gBACX,IAAI,WAAW,GAAG,KAAK,CAAC;gBACxB,KAAK,MAAM,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAC;oBACnC,IAAI,SAAS,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,SAAS,CAAC,MAAM,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC;wBACnE,MAAM,IAAI,GAAG,SAAS,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;wBAChD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC;wBAC1C,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC,CAAC;wBAC9C,WAAW,GAAG,IAAI,CAAC;wBACnB,MAAM;oBACR,CAAC;gBACH,CAAC;gBAED,IAAI,CAAC,WAAW,EAAE,CAAC;oBACjB,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC;gBACjD,CAAC;YACH,CAAC;QACH,CAAC;QAED,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC1E,CAAC;IAED,YAAY,CAAC,IAAY;QACvB,cAAc;QACd,OAAO,IAAI,CAAC,KAAK,CAAC,mCAAmC,CAAC;aACnD,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC3C,CAAC;IAED,UAAU,CAAC,IAAY;QACrB,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC5D,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;IAC9C,CAAC;CACF;AAED;;;GAGG;AACH,MAAM,OAAO,eAAe;IAC1B,QAAQ,GAAa,IAAI,CAAC;IAE1B,0BAA0B;IAClB,QAAQ,GAAG,CAAC,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC;IAEvF,aAAa;IACL,SAAS,GAAG,IAAI,GAAG,CAAC;QAC1B,KAAK,EAAE,GAAG,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO;QACnE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO;QACnE,QAAQ,EAAE,KAAK,EAAE,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;QAClE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,SAAS,EAAE,SAAS,EAAE,MAAM,EAAE,SAAS;QAC/D,QAAQ,EAAE,QAAQ,EAAE,OAAO,EAAE,OAAO,EAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI;QACjE,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,SAAS;QACrE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK;KACvD,CAAC,CAAC;IAEH,QAAQ,CAAC,IAAY;QACnB,MAAM,MAAM,GAAY,EAAE,CAAC;QAC3B,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;QAE9C,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,IAAI,CAAC,IAAI;gBAAE,SAAS;YAEpB,qBAAqB;YACrB,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC;YAE7C,IAAI,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC;gBAClC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC,CAAC;YACrD,CAAC;iBAAM,CAAC;gBACN,qBAAqB;gBACrB,IAAI,WAAW,GAAG,KAAK,CAAC;gBACxB,KAAK,MAAM,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAC;oBACnC,IAAI,SAAS,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,SAAS,CAAC,MAAM,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;wBACvE,MAAM,IAAI,GAAG,SAAS,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;wBAChD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;wBACvD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC,CAAC;wBAC9C,WAAW,GAAG,IAAI,CAAC;wBACnB,MAAM;oBACR,CAAC;gBACH,CAAC;gBAED,IAAI,CAAC,WAAW,EAAE,CAAC;oBACjB,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC;gBACjD,CAAC;YACH,CAAC;QACH,CAAC;QAED,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,KAAK,CAAC,eAAe,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACtE,CAAC;IAED,YAAY,CAAC,IAAY;QACvB,OAAO,IAAI,CAAC,KAAK,CAAC,qDAAqD,CAAC;aACrE,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC3D,CAAC;IAED,UAAU,CAAC,IAAY;QACrB,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC5D,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;IAC9C,CAAC;CACF;AAED;;;GAGG;AACH,MAAM,OAAO,gBAAgB;IAC3B,QAAQ,GAAa,IAAI,CAAC;IAE1B,iBAAiB;IACT,aAAa,GAAG,iBAAiB,CAAC;IAClC,aAAa,GAAG,iBAAiB,CAAC;IAClC,UAAU,GAAG,iBAAiB,CAAC;IAEvC,QAAQ,CAAC,IAAY;QACnB,MAAM,MAAM,GAAY,EAAE,CAAC;QAC3B,IAAI,YAAY,GAAG,EAAE,CAAC;QACtB,IAAI,WAAW,GAAc,MAAM,CAAC;QAEpC,KAAK,MAAM,IAAI,IAAI,IAAI,EAAE,CAAC;YACxB,IAAI,QAAQ,GAAc,MAAM,CAAC;YAEjC,IAAI,IAAI,CAAC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;gBAClC,QAAQ,GAAG,UAAU,CAAC;YACxB,CAAC;iBAAM,IAAI,IAAI,CAAC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;gBACzC,QAAQ,GAAG,UAAU,CAAC;YACxB,CAAC;iBAAM,IAAI,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;gBACtC,QAAQ,GAAG,OAAO,CAAC;YACrB,CAAC;iBAAM,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;gBAC3B,QAAQ,GAAG,YAAY,CAAC;YAC1B,CAAC;iBAAM,IAAI,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;gBAC/B,QAAQ,GAAG,aAAa,CAAC;YAC3B,CAAC;YAED,IAAI,QAAQ,KAAK,WAAW,IAAI,YAAY,EAAE,CAAC;gBAC7C,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,YAAY,EAAE,IAAI,EAAE,WAAW,EAAE,CAAC,CAAC;gBACvD,YAAY,GAAG,EAAE,CAAC;YACpB,CAAC;YAED,IAAI,QAAQ,KAAK,YAAY,EAAE,CAAC;gBAC9B,YAAY,IAAI,IAAI,CAAC;YACvB,CAAC;YACD,WAAW,GAAG,QAAQ,CAAC;QACzB,CAAC;QAED,IAAI,YAAY,EAAE,CAAC;YACjB,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,YAAY,EAAE,IAAI,EAAE,WAAW,EAAE,CAAC,CAAC;QACzD,CAAC;QAED,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACjE,CAAC;IAED,YAAY,CAAC,IAAY;QACvB,OAAO,IAAI,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAChE,CAAC;IAED,UAAU,CAAC,IAAY;QACrB,qEAAqE;QACrE,MAAM,MAAM,GAAG,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;QACnC,OAAO,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,aAAa,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;IACvE,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;IAC9C,CAAC;CACF;AAED;;;GAGG;AACH,MAAM,OAAO,eAAe;IAC1B,QAAQ,GAAa,IAAI,CAAC;IAE1B,0CAA0C;IAClC,SAAS,GAAG,IAAI,GAAG,CAAC;QAC1B,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG;QAChD,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG;QAChD,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG;KACjD,CAAC,CAAC;IAEH,QAAQ,CAAC,IAAY;QACnB,MAAM,MAAM,GAAY,EAAE,CAAC;QAE3B,KAAK,MAAM,IAAI,IAAI,IAAI,EAAE,CAAC;YACxB,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC;gBAAE,SAAS;YAC9B,IAAI,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;gBACzB,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,aAAa,EAAE,CAAC,CAAC;YACnD,CAAC;iBAAM,IAAI,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC;gBACpC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC,CAAC;YAChD,CAAC;iBAAM,CAAC;gBACN,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC,CAAC;YAC7C,CAAC;QACH,CAAC;QAED,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,OAAO,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACjE,CAAC;IAED,YAAY,CAAC,IAAY;QACvB,OAAO,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACjE,CAAC;IAED,UAAU,CAAC,IAAY;QACrB,oCAAoC;QACpC,MAAM,KAAK,GAAG,CAAC,GAAG,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;QAC1E,MAAM,KAAK,GAAa,EAAE,CAAC;QAE3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;YACzC,IAAI,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC;gBACzB,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;YACtC,CAAC;iBAAM,CAAC;gBACN,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;YACvB,CAAC;QACH,CAAC;QAED,OAAO,KAAK,CAAC;IACf,CAAC;IAED,cAAc,CAAC,IAAY;QACzB,0CAA0C;QAC1C,OAAO,CAAC,GAAG,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IAC9C,CAAC;CACF;AAED;;GAEG;AACH,MAAM,UAAU,cAAc,CAAC,IAAY;IACzC,wBAAwB;IACxB,IAAI,MAAM,GAAG,CAAC,EAAE,QAAQ,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,EAAE,KAAK,GAAG,CAAC,CAAC;IAElE,KAAK,MAAM,IAAI,IAAI,IAAI,EAAE,CAAC;QACxB,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC;YAAE,SAAS;QAC9B,KAAK,EAAE,CAAC;QAER,gBAAgB;QAChB,IAAI,2CAA2C,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YAC3D,MAAM,EAAE,CAAC;QACX,CAAC;QACD,gCAAgC;aAC3B,IAAI,8BAA8B,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YACnD,QAAQ,EAAE,CAAC;QACb,CAAC;QACD,6DAA6D;aACxD,IAAI,iBAAiB,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YACtC,OAAO,EAAE,CAAC,CAAE,4BAA4B;QAC1C,CAAC;QACD,QAAQ;aACH,IAAI,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YAC/B,OAAO,EAAE,CAAC;QACZ,CAAC;IACH,CAAC;IAED,IAAI,KAAK,KAAK,CAAC;QAAE,OAAO,IAAI,CAAC;IAE7B,MAAM,OAAO,GAAG,MAAM,GAAG,KAAK,CAAC;IAC/B,MAAM,OAAO,GAAG,QAAQ,GAAG,KAAK,CAAC;IACjC,MAAM,OAAO,GAAG,OAAO,GAAG,KAAK,CAAC;IAChC,MAAM,OAAO,GAAG,OAAO,GAAG,KAAK,CAAC;IAEhC,qBAAqB;IACrB,IAAI,OAAO,GAAG,GAAG;QAAE,OAAO,IAAI,CAAC;IAC/B,IAAI,OAAO,GAAG,GAAG;QAAE,OAAO,IAAI,CAAC,CAAE,+CAA+C;IAChF,IAAI,OAAO,GAAG,GAAG;QAAE,OAAO,IAAI,CAAC;IAC/B,IAAI,OAAO,GAAG,GAAG;QAAE,OAAO,IAAI,CAAC;IAE/B,mBAAmB;IACnB,IAAI,OAAO,GAAG,OAAO,IAAI,OAAO,GAAG,OAAO;QAAE,OAAO,IAAI,CAAC;IACxD,IAAI,OAAO,GAAG,OAAO;QAAE,OAAO,IAAI,CAAC;IACnC,IAAI,OAAO,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC;IAE7B,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,WAAW,CAAC,QAA2B,EAAE,IAAa;IACpE,IAAI,IAAI,GAAG,QAAQ,CAAC;IAEpB,IAAI,QAAQ,KAAK,MAAM,IAAI,IAAI,EAAE,CAAC;QAChC,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;IAC9B,CAAC;IAED,QAAQ,IAAI,EAAE,CAAC;QACb,KAAK,IAAI;YACP,OAAO,IAAI,cAAc,EAAE,CAAC;QAC9B,KAAK,IAAI;YACP,OAAO,IAAI,gBAAgB,EAAE,CAAC;QAChC,KAAK,IAAI;YACP,OAAO,IAAI,eAAe,EAAE,CAAC;QAC/B,KAAK,IAAI,CAAC;QACV;YACE,OAAO,IAAI,eAAe,EAAE,CAAC;IACjC,CAAC;AACH,CAAC;AAED;;GAEG;AACH,MAAM,OAAO,sBAAsB;IACzB,SAAS,GAAoC,IAAI,GAAG,EAAE,CAAC;IAE/D;QACE,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,IAAI,EAAE,IAAI,cAAc,EAAE,CAAC,CAAC;QAC/C,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,IAAI,EAAE,IAAI,eAAe,EAAE,CAAC,CAAC;QAChD,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,IAAI,EAAE,IAAI,gBAAgB,EAAE,CAAC,CAAC;QACjD,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,IAAI,EAAE,IAAI,eAAe,EAAE,CAAC,CAAC;IAClD,CAAC;IAED;;OAEG;IACH,QAAQ,CAAC,IAAY;QACnB,MAAM,QAAQ,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;QACtC,MAAM,QAAQ,GAAG,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,QAAQ,CAAC,IAAI,IAAI,eAAe,EAAE,CAAC;QACvE,OAAO;YACL,QAAQ;YACR,MAAM,EAAE,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC;SAChC,CAAC;IACJ,CAAC;IAED;;OAEG;IACH,OAAO,CAAC,IAAY;QAQlB,MAAM,QAAQ,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;QACtC,MAAM,QAAQ,GAAG,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,QAAQ,CAAC,IAAI,IAAI,eAAe,EAAE,CAAC;QAEvE,OAAO;YACL,QAAQ;YACR,SAAS,EAAE,QAAQ,CAAC,cAAc,CAAC,IAAI,CAAC;YACxC,OAAO,EAAE,QAAQ,CAAC,YAAY,CAAC,IAAI,CAAC;YACpC,KAAK,EAAE,QAAQ,CAAC,UAAU,CAAC,IAAI,CAAC;YAChC,SAAS,EAAE,QAAQ,CAAC,cAAc,CAAC,IAAI,CAAC;YACxC,MAAM,EAAE,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC;SAChC,CAAC;IACJ,CAAC;CACF;AAED,mBAAmB;AACnB,MAAM,CAAC,MAAM,kBAAkB,GAAG,IAAI,sBAAsB,EAAE,CAAC"}

package/dist/core/local-embedding.d.ts

@@ -1,109 +0,0 @@
-/**
- * Local Embedding Service
- * 로컬 초소형 모델을 사용한 임베딩 생성
- *
- * @module core/local-embedding
- */
-/**
- * 임베딩 결과
- */
-interface EmbeddingResult {
-    text: string;
-    embedding: number[];
-    model: string;
-    processingTime: number;
-}
-/**
- * 배치 임베딩 결과
- */
-interface BatchEmbeddingResult {
-    results: EmbeddingResult[];
-    totalTime: number;
-    avgTime: number;
-}
-/**
- * 유사도 결과
- */
-interface SimilarityResult {
-    text1: string;
-    text2: string;
-    similarity: number;
-    method: 'cosine' | 'dot' | 'euclidean';
-}
-/**
- * Local Embedding Service
- */
-export declare class LocalEmbeddingService {
-    private extractor;
-    private modelName;
-    private isLoading;
-    private loadPromise;
-    private cache;
-    private cacheMaxSize;
-    private stats;
-    constructor(modelName?: string);
-    /**
-     * 모델 로드
-     */
-    init(): Promise<void>;
-    /**
-     * 단일 텍스트 임베딩
-     */
-    embed(text: string): Promise<EmbeddingResult>;
-    /**
-     * 배치 임베딩
-     */
-    embedBatch(texts: string[]): Promise<BatchEmbeddingResult>;
-    /**
-     * 두 텍스트 간 유사도 계산
-     */
-    similarity(text1: string, text2: string, method?: 'cosine' | 'dot' | 'euclidean'): Promise<SimilarityResult>;
-    /**
-     * 쿼리와 가장 유사한 텍스트 찾기
-     */
-    findMostSimilar(query: string, candidates: string[], topK?: number): Promise<Array<{
-        text: string;
-        similarity: number;
-    }>>;
-    /**
-     * 코사인 유사도
-     */
-    private cosineSimilarity;
-    /**
-     * 내적
-     */
-    private dotProduct;
-    /**
-     * 유클리드 거리
-     */
-    private euclideanDistance;
-    /**
-     * 캐시에 추가
-     */
-    private addToCache;
-    /**
-     * 통계 반환
-     */
-    getStats(): {
-        cacheSize: number;
-        cacheHitRate: number;
-        avgEmbeddingTime: number;
-        model: string;
-        isLoaded: boolean;
-        totalEmbeddings: number;
-        cacheHits: number;
-        totalTime: number;
-        modelLoadTime: number;
-    };
-    /**
-     * 캐시 클리어
-     */
-    clearCache(): void;
-    /**
-     * 임베딩 차원 반환
-     */
-    getDimension(): number;
-}
-export declare const localEmbedding: LocalEmbeddingService;
-export {};
-//# sourceMappingURL=local-embedding.d.ts.map

package/dist/core/local-embedding.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"local-embedding.d.ts","sourceRoot":"","sources":["../../src/core/local-embedding.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AAYH;;GAEG;AACH,UAAU,eAAe;IACvB,IAAI,EAAE,MAAM,CAAC;IACb,SAAS,EAAE,MAAM,EAAE,CAAC;IACpB,KAAK,EAAE,MAAM,CAAC;IACd,cAAc,EAAE,MAAM,CAAC;CACxB;AAED;;GAEG;AACH,UAAU,oBAAoB;IAC5B,OAAO,EAAE,eAAe,EAAE,CAAC;IAC3B,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;CACjB;AAED;;GAEG;AACH,UAAU,gBAAgB;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,UAAU,EAAE,MAAM,CAAC;IACnB,MAAM,EAAE,QAAQ,GAAG,KAAK,GAAG,WAAW,CAAC;CACxC;AAED;;GAEG;AACH,qBAAa,qBAAqB;IAChC,OAAO,CAAC,SAAS,CAAiC;IAClD,OAAO,CAAC,SAAS,CAAS;IAC1B,OAAO,CAAC,SAAS,CAAkB;IACnC,OAAO,CAAC,WAAW,CAA8B;IAGjD,OAAO,CAAC,KAAK,CAAoC;IACjD,OAAO,CAAC,YAAY,CAAiB;IAGrC,OAAO,CAAC,KAAK,CAKX;gBAEU,SAAS,GAAE,MAAkC;IAIzD;;OAEG;IACG,IAAI,IAAI,OAAO,CAAC,IAAI,CAAC;IA0B3B;;OAEG;IACG,KAAK,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,eAAe,CAAC;IAwCnD;;OAEG;IACG,UAAU,CAAC,KAAK,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,oBAAoB,CAAC;IAqBhE;;OAEG;IACG,UAAU,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,GAAE,QAAQ,GAAG,KAAK,GAAG,WAAsB,GAAG,OAAO,CAAC,gBAAgB,CAAC;IAuB5H;;OAEG;IACG,eAAe,CAAC,KAAK,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,EAAE,EAAE,IAAI,GAAE,MAAU,GAAG,OAAO,CAAC,KAAK,CAAC;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IAclI;;OAEG;IACH,OAAO,CAAC,gBAAgB;IAcxB;;OAEG;IACH,OAAO,CAAC,UAAU;IAIlB;;OAEG;IACH,OAAO,CAAC,iBAAiB;IAIzB;;OAEG;IACH,OAAO,CAAC,UAAU;IASlB;;OAEG;IACH,QAAQ;;;;;;;;;;;IAeR;;OAEG;IACH,UAAU,IAAI,IAAI;IAIlB;;OAEG;IACH,YAAY,IAAI,MAAM;CAGvB;AAGD,eAAO,MAAM,cAAc,uBAA8B,CAAC"}