Outline Wiki MCP Server

/** * Embedding Service * * Converts text to vector embeddings using OpenAI API */ import OpenAI from 'openai'; import { EMBEDDING, ERROR_MESSAGES } from './constants.js'; import type { IEmbeddingService, EmbeddingConfig } from './types.js'; export class EmbeddingService implements IEmbeddingService { private client: OpenAI | null = null; private readonly model: string; private readonly dimensions: number; constructor(apiKey?: string, config: EmbeddingConfig = {}) { this.model = config.model ?? EMBEDDING.MODEL; this.dimensions = config.dimensions ?? EMBEDDING.DIMENSIONS; if (apiKey) { this.client = new OpenAI({ apiKey }); } } isEnabled(): boolean { return this.client !== null; } getDimensions(): number { return this.dimensions; } async getEmbedding(text: string): Promise<number[]> { if (!this.client) { throw new Error(ERROR_MESSAGES.OPENAI_NOT_CONFIGURED); } // Clean text for better embedding quality const cleanText = text.replace(/\n+/g, ' ').trim(); if (!cleanText) { return Array(this.dimensions).fill(0); } const response = await this.client.embeddings.create({ model: this.model, input: cleanText, }); return response.data[0].embedding; } async getEmbeddings(texts: string[]): Promise<number[][]> { if (!this.client) { throw new Error(ERROR_MESSAGES.OPENAI_NOT_CONFIGURED); } const cleanTexts = texts.map((t) => t.replace(/\n+/g, ' ').trim()); const response = await this.client.embeddings.create({ model: this.model, input: cleanTexts, }); return response.data.map((d) => d.embedding); } }