Skip to content

TEST_REPORT.md

Latest commit

 

History

History
244 lines (178 loc) · 7.39 KB

TEST_REPORT.md

File metadata and controls

244 lines (178 loc) · 7.39 KB

AskOnce 运行测试报告 / Test Report

测试日期 / Test Date: 2026-03-07 状态 / Status: ✅ 全部测试通过 / All Tests Passed!

测试环境 / Test Environment

项目 / Item 版本/值 / Version/Value
Node.js v22.22.0
npm 10.9.4
操作系统 / OS Linux 5.15.77-amd64-desktop
API Key sk-xxxxxxxxxxxxxxxx (DeepSeek)

运行步骤 / Running Steps

1. 创建 .env 配置文件 / Create .env Configuration File

# 在项目根目录创建 .env 文件 / Create .env file in project root
DEEPSEEK_API_KEY=sk-xxxxxxxxxxxxxxxx
PORT=8000
MAX_CONCURRENT=10
DEFAULT_TIMEOUT=60000

2. 安装依赖 / Install Dependencies

# 前端依赖 / Frontend dependencies
cd frontend
npm install

# 后端依赖 / Backend Dependencies
cd ../backend/node
npm install

3. 修复 .env 加载路径 / Fix .env Loading Path

修改 backend/node/src/index.ts，添加从项目根目录加载 .env：

Modify backend/node/src/index.ts to load .env from project root:

import path from 'path';
import { fileURLToPath } from 'url';

const __filename = fileURLToPath(import.meta.url);
const __dirname = path.dirname(__filename);
const rootDir = path.resolve(__dirname, '../../..');

dotenv.config({ path: path.join(rootDir, '.env') });

4. 启动后端服务 / Start Backend Service

cd backend/node
npm run dev

输出 / Output:

╔═══════════════════════════════════════════════════════════╗
║   AskOnce API Server                                      ║
║   Server running at: http://localhost:8000              ║
║   WebSocket at: ws://localhost:8000/ws                  ║
╚═══════════════════════════════════════════════════════════╝

5. 测试 API / Test API

# 健康检查 / Health check
curl http://localhost:8000/api/health
# 输出 / Output: {"status":"healthy","timestamp":"2026-03-07T12:26:10.674Z"}

# 单模型查询 / Single model query
curl -X POST http://localhost:8000/api/query \
  -H "Content-Type: application/json" \
  -d '{"prompt": "你好，请用一句话介绍自己", "model_ids": ["deepseek-v3"]}'

测试结果 / Test Results

测试项 / Test Item 结果 / Result 说明 / Description
后端启动 / Backend Startup ✅ 成功 / Success 监听端口 8000 / Listening on port 8000
健康检查 / Health Check ✅ 正常 / Normal 返回 healthy 状态 / Returns healthy status
获取模型列表 / Get Model List ✅ 正常 / Normal 返回 14 个模型 / Returns 14 models
前端启动 / Frontend Startup ✅ 成功 / Success 监听端口 3002 / Listening on port 3002
单模型查询 / Single Model Query ✅ 成功 / Success DeepSeek V3 正常响应 / DeepSeek V3 responded normally
多模型并发查询 / Multi-model Query ✅ 成功 / Success V3 和 R1 同时响应 / V3 and R1 responded simultaneously

API 测试详情 / API Test Details

测试 1: 单模型查询 / Test 1: Single Model Query

curl -X POST http://localhost:8000/api/query \
  -H "Content-Type: application/json" \
  -d '{"prompt": "你好，请用一句话介绍自己", "model_ids": ["deepseek-v3"]}'

结果 / Result:

{
  "request_id": "req_1772886383770",
  "status": "completed",
  "results": [{
    "model_id": "deepseek-v3",
    "status": "completed",
    "content": "你好，我是DeepSeek，由深度求索公司创造的AI助手，乐于为你提供解答和帮助！😊",
    "tokens_used": 36,
    "duration": 2025
  }]
}

耗时 / Duration: 2.025 秒 / seconds

测试 2: 多模型并发查询 / Test 2: Multi-model Concurrent Query

curl -X POST http://localhost:8000/api/query \
  -H "Content-Type: application/json" \
  -d '{"prompt": "什么是量子计算？用一句话解释", "model_ids": ["deepseek-v3", "deepseek-r1"]}'

结果 / Result:

{
  "request_id": "req_1772886396354",
  "status": "completed",
  "results": [
    {
      "model_id": "deepseek-v3",
      "status": "completed",
      "content": "量子计算是利用量子比特的叠加与纠缠特性，在特定问题上实现远超经典计算机算力的新型计算模式。",
      "tokens_used": 35,
      "duration": 1828
    },
    {
      "model_id": "deepseek-r1",
      "status": "completed",
      "content": "量子计算是一种利用量子叠加与纠缠等特性，通过量子比特实现指数级加速信息处理的新型计算方式。",
      "tokens_used": 55,
      "duration": 2637
    }
  ]
}

耗时 / Duration: 2.637 秒 (两个模型并发执行 / Two models executed concurrently)

测试 3: 前端页面 / Test 3: Frontend Page

curl http://localhost:3002

结果 / Result: HTML 页面正常返回，包含 AskOnce 标题 / HTML page returned normally with AskOnce title

已知问题 (已修复) / Known Issues (Fixed)

1. ✅ 已修复: API 认证失败 / Fixed: API Authentication Failed

问题 / Issue: 使用示例 API Key 导致 401 认证错误 / Sample API Key caused 401 authentication error

解决 / Solution: 配置真实的 DeepSeek API Key / Configure real DeepSeek API Key

2. ✅ 已修复: .env 加载路径 / Fixed: .env Loading Path

问题 / Issue: .env 文件在项目根目录，但后端从 backend/node 运行 / .env in project root, but backend runs from backend/node

解决 / Solution: 修改 dotenv.config() 加载路径 / Modify dotenv.config() loading path

启动命令汇总 / Startup Commands Summary

# 1. 配置 API Key (已配置) / Configure API Key (Configured)
DEEPSEEK_API_KEY=sk-xxxxxxxxxxxxxxxx

# 2. 安装依赖 (已完成) / Install dependencies (Completed)
cd frontend && npm install
cd ../backend/node && npm install

# 3. 启动服务 (两个终端) / Start services (Two terminals)

# 终端1: 后端 / Terminal 1: Backend
cd backend/node && npm run dev

# 终端2: 前端 / Terminal 2: Frontend
cd frontend && npm run dev

# 4. 访问 / Access
# 前端 / Frontend: http://localhost:3000 (或 3001/3002 如果端口被占用 / or 3001/3002 if port is occupied)
# 后端 / Backend: http://localhost:8000

功能验证 / Feature Verification

功能 / Feature 状态 / Status 备注 / Notes
并发查询 / Concurrent Query ✅ 两个 DeepSeek 模型同时响应 / Two DeepSeek models responded simultaneously
流式响应 / Streaming Response 🔄 代码已实现，需前端测试 / Code implemented, needs frontend testing
AI 评审 / AI Judge 🔄 需要更多模型 / Needs more models
多视图展示 / Multi-view Display 🔄 需要前端界面测试 / Needs frontend UI testing

结论 / Conclusion

✅ AskOnce 项目运行成功！/ AskOnce Project Running Successfully!

  • 后端 API 正常运行 / Backend API running normally
  • DeepSeek 模型调用成功 / DeepSeek model called successfully
  • 多模型并发查询工作正常 / Multi-model concurrent query working normally
  • 前端页面可访问 / Frontend page accessible

报告版本 / Report Version: v1.2 更新日期 / Updated: 2026-03-07 作者 / Author: SeekSage