MCP Advisor

/** * 向量工具函数测试 * 测试向量操作相关的工具函数 */ import { describe, test, expect, vi } from 'vitest'; import { normalizeVector, cosineSimilarity } from '../../../utils/vectorUtils.js'; // 实现测试中需要的 createZeroVector 函数 function createZeroVector(dimension: number): number[] { return new Array(dimension).fill(0); } describe('向量工具函数', () => { describe('normalizeVector', () => { test('应该正确归一化非零向量', () => { const vector = [3, 4, 0]; const normalized = normalizeVector(vector); // 验证向量长度为 1 const magnitude = Math.sqrt( normalized.reduce((sum: number, val: number) => sum + val * val, 0) ); expect(magnitude).toBeCloseTo(1); // 验证方向不变 expect(normalized[0]).toBeCloseTo(0.6); expect(normalized[1]).toBeCloseTo(0.8); expect(normalized[2]).toBeCloseTo(0); }); test('应该处理零向量', () => { const vector = [0, 0, 0]; const normalized = normalizeVector(vector); // 零向量归一化后仍为零向量 expect(normalized).toEqual([0, 0, 0]); }); test.todo('应该处理单元素向量'); test.todo('应该处理非常大的向量值'); test.todo('应该处理非常小的向量值'); }); describe('cosineSimilarity', () => { test('相同向量的相似度应为 1', () => { const v1 = [1, 2, 3]; const v2 = [1, 2, 3]; const similarity = cosineSimilarity(v1, v2); expect(similarity).toBeCloseTo(1); }); test('正交向量的相似度应为 0', () => { const v1 = [1, 0, 0]; const v2 = [0, 1, 0]; const similarity = cosineSimilarity(v1, v2); expect(similarity).toBeCloseTo(0); }); test('方向相反的向量相似度应为 -1', () => { const v1 = [1, 2, 3]; const v2 = [-1, -2, -3]; const similarity = cosineSimilarity(v1, v2); expect(similarity).toBeCloseTo(-1); }); test.todo('应该处理零向量'); test.todo('应该处理不同维度的向量'); }); describe('createZeroVector', () => { test('应该创建指定维度的零向量', () => { const dim = 5; const zeroVector = createZeroVector(dim); expect(zeroVector.length).toBe(dim); expect(zeroVector.every(val => val === 0)).toBe(true); }); test.todo('应该处理维度为 0 的情况'); test.todo('应该处理非常大的维度'); }); });