Naver Search MCP Server

// Import JSON data - bundle-safe approach import { existsSync, readFileSync } from 'fs'; import { join } from 'path'; // 메모리 누수 방지를 위한 지연 로딩 캐시 let categoriesCache: any[] | null = null; // Load categories data with fallback paths for different environments function getCategoriesData(): any[] { // 캐시된 데이터가 있으면 반환 (메모리 효율성) if (categoriesCache !== null) { return categoriesCache; } try { // Try bundled data path first (dist/data) const bundledPath = join(process.cwd(), 'dist', 'data', 'categories.json'); if (existsSync(bundledPath)) { categoriesCache = JSON.parse(readFileSync(bundledPath, 'utf8') as string); return categoriesCache!; } // Fallback to source data path const sourcePath = join(process.cwd(), 'data', 'categories.json'); if (existsSync(sourcePath)) { categoriesCache = JSON.parse(readFileSync(sourcePath, 'utf8') as string); return categoriesCache!; } throw new Error('카테고리 데이터 파일을 찾을 수 없습니다'); } catch (error) { console.error('카테고리 데이터 로딩 실패:', error); return []; } } /** * 캐시 정리 함수 (메모리 누수 방지) */ export function clearCategoriesCache(): void { categoriesCache = null; } // Category data structure interface CategoryData { code: string; level1: string; // 대분류 level2: string; // 중분류 level3: string; // 소분류 level4: string; // 세분류 } /** * Load category data from bundled JSON file */ async function loadCategoryData(): Promise<CategoryData[]> { // 지연 로딩된 캐시 데이터 사용 const data = getCategoriesData(); console.error(`Loaded ${data.length} categories from bundled JSON data`); return data as CategoryData[]; } /** * Smart category search with fuzzy matching and ranking */ async function smartCategorySearch(query: string, maxResults: number = 10): Promise<Array<CategoryData & { score: number, matchType: string }>> { const categories = await loadCategoryData(); const searchQuery = query.toLowerCase().trim(); const results: Array<CategoryData & { score: number, matchType: string }> = []; for (const category of categories) { let bestScore = 0; let bestMatchType = ''; // Check all levels for matches const levels = [ { value: category.level1, name: '대분류' }, { value: category.level2, name: '중분류' }, { value: category.level3, name: '소분류' }, { value: category.level4, name: '세분류' } ]; for (const level of levels) { if (!level.value) continue; const levelText = level.value.toLowerCase(); let score = 0; let matchType = ''; // Level-based bonus (대분류 우선) let levelBonus = 0; if (level.name === '대분류') levelBonus = 50; else if (level.name === '중분류') levelBonus = 30; else if (level.name === '소분류') levelBonus = 20; else if (level.name === '세분류') levelBonus = 10; // Exact match (highest priority) if (levelText === searchQuery) { score = 100 + levelBonus; matchType = `정확일치(${level.name})`; } // Starts with (high priority) else if (levelText.startsWith(searchQuery)) { score = 80 + levelBonus; matchType = `시작일치(${level.name})`; } // Contains (medium priority) else if (levelText.includes(searchQuery)) { score = 60 + levelBonus; matchType = `포함일치(${level.name})`; } // Fuzzy match for similar terms else { const similarity = calculateSimilarity(searchQuery, levelText); if (similarity > 0.6) { score = Math.floor(similarity * 40) + levelBonus; // 0.6-1.0 -> 24-40 points + level bonus matchType = `유사일치(${level.name})`; } } if (score > bestScore) { bestScore = score; bestMatchType = matchType; } } if (bestScore > 0) { results.push({ ...category, score: bestScore, matchType: bestMatchType }); } } // Sort by score (descending), then by category code (ascending for 대분류 priority) return results .sort((a, b) => { if (b.score !== a.score) { return b.score - a.score; // Higher score first } // If same score, prioritize by category code (대분류 codes are typically smaller) return a.code.localeCompare(b.code); }) .slice(0, maxResults); } /** * Calculate similarity between two strings (simple implementation) */ function calculateSimilarity(str1: string, str2: string): number { const longer = str1.length > str2.length ? str1 : str2; const shorter = str1.length > str2.length ? str2 : str1; if (longer.length === 0) return 1.0; const editDistance = levenshteinDistance(longer, shorter); return (longer.length - editDistance) / longer.length; } /** * Calculate Levenshtein distance between two strings */ function levenshteinDistance(str1: string, str2: string): number { const matrix = Array(str2.length + 1).fill(null).map(() => Array(str1.length + 1).fill(null)); for (let i = 0; i <= str1.length; i++) matrix[0][i] = i; for (let j = 0; j <= str2.length; j++) matrix[j][0] = j; for (let j = 1; j <= str2.length; j++) { for (let i = 1; i <= str1.length; i++) { const substitutionCost = str1[i - 1] === str2[j - 1] ? 0 : 1; matrix[j][i] = Math.min( matrix[j][i - 1] + 1, matrix[j - 1][i] + 1, matrix[j - 1][i - 1] + substitutionCost ); } } return matrix[str2.length][str1.length]; } /** * 카테고리 검색 핸들러 */ export const findCategoryHandler = async ({ query, max_results = 10 }: any) => { try { const results = await smartCategorySearch(query, max_results); if (results.length === 0) { return { message: `"${query}"와 관련된 카테고리를 찾을 수 없습니다. 다른 검색어를 시도해보세요.`, suggestions: ["패션", "화장품", "가구", "스마트폰", "가전제품", "스포츠", "도서", "자동차", "식품", "뷰티"] }; } const responseData: any = { message: `"${query}" 검색 결과 (${results.length}개, 관련도순 정렬)`, total_found: results.length, categories: results.map(cat => ({ code: cat.code, category: [cat.level1, cat.level2, cat.level3, cat.level4].filter(Boolean).join(' > '), match_type: cat.matchType, score: cat.score, levels: { 대분류: cat.level1 || '', 중분류: cat.level2 || '', 소분류: cat.level3 || '', 세분류: cat.level4 || '' } })), next_steps: { trend_analysis: `이제 datalab_shopping_category 도구로 각 카테고리의 트렌드 분석이 가능합니다`, age_analysis: `datalab_shopping_age 도구로 연령별 쇼핑 패턴을 분석할 수 있습니다`, gender_analysis: `datalab_shopping_gender 도구로 성별 쇼핑 패턴을 분석할 수 있습니다`, device_analysis: `datalab_shopping_device 도구로 디바이스별 쇼핑 패턴을 분석할 수 있습니다` } }; return responseData; } catch (error: any) { throw new Error(`카테고리 검색 중 오류 발생: ${error.message}`); } }; export const categoryToolHandlers: Record<string, (args: any) => Promise<any>> = { find_category: findCategoryHandler, };