Design Patterns MCP Server

/** * Strategy Pattern for Embedding Generation * Defines different algorithms for generating text embeddings */ export interface EmbeddingVector { dimensions: number; values: number[]; model: string; normalized: boolean; } export interface EmbeddingStrategy { readonly name: string; readonly dimensions: number; readonly model: string; generateEmbedding(text: string): Promise<EmbeddingVector>; batchGenerateEmbeddings(texts: string[]): Promise<EmbeddingVector[]>; isAvailable(): Promise<boolean>; } /** * Simple Hash-based Embedding Strategy (fallback) */ export class SimpleHashEmbeddingStrategy implements EmbeddingStrategy { readonly name = 'simple-hash'; readonly dimensions = 384; readonly model = 'simplified-hash'; async generateEmbedding(text: string): Promise<EmbeddingVector> { const words = text.toLowerCase().split(/\s+/); const embedding = new Array(this.dimensions).fill(0); // Improved hash-based algorithm for (let i = 0; i < words.length; i++) { const word = words[i]; const wordHash = this.simpleHash(word); for (let j = 0; j < Math.min(word.length, 10); j++) { const charCode = word.charCodeAt(j); const position = (wordHash + j + i * 7) % embedding.length; embedding[position] += (charCode / 255) * 0.5 + Math.sin(wordHash * j) * 0.3; } } // Normalize the embedding const norm = Math.sqrt(embedding.reduce((sum, val) => sum + val * val, 0)); const normalizedValues = embedding.map(val => val / (norm || 1)); return { dimensions: this.dimensions, values: normalizedValues, model: this.model, normalized: true, }; } async batchGenerateEmbeddings(texts: string[]): Promise<EmbeddingVector[]> { return Promise.all(texts.map(text => this.generateEmbedding(text))); } async isAvailable(): Promise<boolean> { return true; // Always available as fallback } private simpleHash(text: string): number { let hash = 0; for (let i = 0; i < text.length; i++) { const char = text.charCodeAt(i); hash = (hash << 5) - hash + char; hash = hash & hash; // Convert to 32-bit integer } return Math.abs(hash); } } /** * Transformers.js Embedding Strategy (primary) */ export class TransformersEmbeddingStrategy implements EmbeddingStrategy { readonly name = 'transformers-js'; readonly dimensions = 384; readonly model = 'all-MiniLM-L6-v2'; private pipeline: any = null; private isInitialized = false; async generateEmbedding(text: string): Promise<EmbeddingVector> { await this.initialize(); if (!this.pipeline) { throw new Error('Transformers.js pipeline not available'); } try { const response = await this.pipeline([text], { pooling: 'mean', normalize: true, }); const data = response.data as unknown; if (!Array.isArray(data) && !(data instanceof Float32Array)) { throw new Error('Invalid embedding data: expected array or Float32Array of numbers'); } const arrayData = Array.isArray(data) ? data : Array.from(data); if (!arrayData.every((v: unknown) => typeof v === 'number')) { throw new Error('Invalid embedding data: expected array of numbers'); } const values = arrayData; return { dimensions: this.dimensions, values, model: this.model, normalized: true, }; } catch (error) { throw new Error( `Embedding generation failed: ${error instanceof Error ? error.message : 'Unknown error'}` ); } } async batchGenerateEmbeddings(texts: string[]): Promise<EmbeddingVector[]> { await this.initialize(); if (!this.pipeline) { throw new Error('Transformers.js pipeline not available'); } try { const response = await this.pipeline(texts, { pooling: 'mean', normalize: true, }); const data = response.data as unknown; if (!Array.isArray(data) && !(data instanceof Float32Array)) { throw new Error('Invalid embedding data: expected array or Float32Array of numbers'); } const embeddingData = Array.isArray(data) ? (data as number[]) : Array.from(data); // Handle batch response const embeddings: EmbeddingVector[] = []; for (let i = 0; i < texts.length; i++) { const values = embeddingData.slice(i * this.dimensions, (i + 1) * this.dimensions); embeddings.push({ dimensions: this.dimensions, values, model: this.model, normalized: true, }); } return embeddings; } catch (error) { throw new Error( `Batch embedding generation failed: ${error instanceof Error ? error.message : 'Unknown error'}` ); } } async isAvailable(): Promise<boolean> { try { await this.initialize(); return this.pipeline !== null; } catch { return false; } } private async initialize(): Promise<void> { if (this.isInitialized) return; try { // Dynamic import to avoid issues if @xenova/transformers is not installed const transformersModule = await import('@xenova/transformers').catch(() => null); if (transformersModule) { this.pipeline = await transformersModule.pipeline( 'feature-extraction', 'Xenova/all-MiniLM-L6-v2' ); } else { throw new Error('Transformers.js module not available'); } this.isInitialized = true; } catch (error) { console.warn('Transformers.js not available, will use fallback strategy'); this.isInitialized = true; // Prevent retries this.pipeline = null; } } } /** * Ollama Embedding Strategy (alternative) */ export class OllamaEmbeddingStrategy implements EmbeddingStrategy { readonly name = 'ollama'; readonly dimensions = 384; readonly model = 'all-minilm:l6-v2'; private baseUrl: string; constructor(baseUrl = 'http://localhost:11434') { this.baseUrl = baseUrl; } async generateEmbedding(text: string): Promise<EmbeddingVector> { try { const response = await fetch(`${this.baseUrl}/api/embeddings`, { method: 'POST', headers: { 'Content-Type': 'application/json', }, body: JSON.stringify({ model: this.model, prompt: text, }), }); if (!response.ok) { throw new Error(`Ollama API error: ${response.statusText}`); } const data = await response.json(); return { dimensions: data.embedding.length, values: data.embedding, model: this.model, normalized: false, // Ollama may not normalize by default }; } catch (error) { throw new Error( `Ollama embedding generation failed: ${error instanceof Error ? error.message : 'Unknown error'}` ); } } async batchGenerateEmbeddings(texts: string[]): Promise<EmbeddingVector[]> { // Ollama doesn't support batch processing, so we process sequentially const embeddings: EmbeddingVector[] = []; for (const text of texts) { const embedding = await this.generateEmbedding(text); embeddings.push(embedding); } return embeddings; } async isAvailable(): Promise<boolean> { try { const response = await fetch(`${this.baseUrl}/api/tags`, { method: 'GET', signal: AbortSignal.timeout(5000), // 5 second timeout }); if (!response.ok) return false; const data = await response.json(); return ( data.models?.some( (model: any) => model.name.includes('all-minilm') || model.name.includes('embedding') ) || false ); } catch { return false; } } }