refactor services to use typescript

server/src/services/retrievalService.ts

@@ -0,0 +1,131 @@
+'use strict';
+
+import logger from '../logger.js';
+import * as vectorStore from '../stores/vectorStore.js';
+import * as notebookStore from '../stores/notebookStore.js';
+
+const VOYAGE_API_URL = 'https://api.voyageai.com/v1';
+const EMBED_MODEL = 'voyage-3';
+const RERANK_MODEL = 'rerank-2';
+
+function getApiKey(): string {
+  const key = process.env.VOYAGE_API_KEY;
+  if (!key) throw new Error('VOYAGE_API_KEY is not set');
+  return key;
+}
+
+export interface TextChunk {
+  id: string;
+  text: string;
+  sourceId: string;
+  index: number;
+}
+
+interface VoyageEmbeddingResponse {
+  data: Array<{ embedding: number[] }>;
+}
+
+interface VoyageRerankResponse {
+  data: Array<{ index: number; relevance_score: number }>;
+}
+
+export interface ScoredChunk {
+  chunk: TextChunk;
+  score: number;
+}
+
+export interface RankedChunk {
+  chunk: TextChunk;
+  relevanceScore: number;
+}
+
+export async function embedTexts(
+  texts: string[],
+  inputType: 'document' | 'query' = 'document'
+): Promise<number[][]> {
+  const res = await fetch(`${VOYAGE_API_URL}/embeddings`, {
+    method: 'POST',
+    headers: {
+      'Content-Type': 'application/json',
+      Authorization: `Bearer ${getApiKey()}`,
+    },
+    body: JSON.stringify({
+      model: EMBED_MODEL,
+      input: texts,
+      input_type: inputType,
+    }),
+  });
+
+  if (!res.ok) {
+    const body = await res.text();
+    throw new Error(`Voyage embed failed (${res.status}): ${body}`);
+  }
+
+  const json = (await res.json()) as VoyageEmbeddingResponse;
+  return json.data.map((d) => d.embedding);
+}
+
+export function storeChunkEmbeddings(chunks: TextChunk[], embeddings: number[][]): void {
+  for (let i = 0; i < chunks.length; i++) {
+    vectorStore.storeVector(chunks[i].id, embeddings[i]);
+  }
+  logger.debug({ count: chunks.length }, 'stored chunk embeddings');
+}
+
+export function search(queryEmbedding: number[], notebookId: string, topK = 10): ScoredChunk[] {
+  const chunks = notebookStore.getChunksForNotebook(notebookId) as TextChunk[];
+  if (chunks.length === 0) return [];
+
+  const scored: ScoredChunk[] = [];
+  for (const chunk of chunks) {
+    const vec = vectorStore.getVector(chunk.id);
+    if (!vec) continue;
+    scored.push({ chunk, score: cosineSimilarity(queryEmbedding, vec) });
+  }
+
+  scored.sort((a, b) => b.score - a.score);
+  return scored.slice(0, topK);
+}
+
+export async function rerank(query: string, chunks: TextChunk[]): Promise<RankedChunk[]> {
+  if (chunks.length === 0) return [];
+
+  const res = await fetch(`${VOYAGE_API_URL}/rerank`, {
+    method: 'POST',
+    headers: {
+      'Content-Type': 'application/json',
+      Authorization: `Bearer ${getApiKey()}`,
+    },
+    body: JSON.stringify({
+      model: RERANK_MODEL,
+      query,
+      documents: chunks.map((c) => c.text),
+    }),
+  });
+
+  if (!res.ok) {
+    const body = await res.text();
+    throw new Error(`Voyage rerank failed (${res.status}): ${body}`);
+  }
+
+  const json = (await res.json()) as VoyageRerankResponse;
+  return json.data.map((item) => ({
+    chunk: chunks[item.index],
+    relevanceScore: item.relevance_score,
+  }));
+}
+
+export function cosineSimilarity(a: number[], b: number[]): number {
+  let dot = 0;
+  let normA = 0;
+  let normB = 0;
+
+  for (let i = 0; i < a.length; i++) {
+    dot += a[i] * b[i];
+    normA += a[i] * a[i];
+    normB += b[i] * b[i];
+  }
+
+  const denom = Math.sqrt(normA) * Math.sqrt(normB);
+  return denom === 0 ? 0 : dot / denom;
+}