How liyupi/yudada works
Project type and purpose
• From the file tree and README, this is a full‑stack “AI 答题应用平台” (AI‑powered quiz application platform) comprising:
• A Taro+React “MBTI 性格测试小程序” mini‑app (mbti-test-mini)
• A Vue 3 web frontend (yudada-frontend)
• A Java Spring Boot backend (yudada-backend)
• Purpose: let users create, publish, share and answer quiz apps, optionally backed by AI generation/scoring; admin can manage apps, users, and view analytics.
File structure
• Top‑level modules:
• mbti-test-mini: Taro cross‑platform微信小程序, with src/components, pages, utils, data, configs
• yudada-frontend: Vue 3 SPA, with src/api, components, views, router, store (Pinia), OpenAPI codegen
• yudada-backend: Spring Boot monolith with layered package structure (controller, service, manager, mapper, entity, dto, config, utils), plus SQL scripts, Dockerfile, tests
• Config/build files:
• Frontend: vue.config.js, babel.config.js, Dockerfile, nginx.conf
• Backend: pom.xml, application.yml (dev/prod), Dockerfile, Maven wrapper
• Patterns:
• Backend: MVC (controllers → services → mappers → DB), AOP interceptors, annotation‑driven auth, RxJava reactive flows, SSE streams, caching layers, sharding
• Frontend: MVVM (Vue components + Pinia store), OpenAPI‑driven API client, role-based access control
README for additional insights
• Architecture section shows:
• Tech stack summary: MySQL, Redis, COS, ChatGLM AI, SSE, RxJava, ShardingSphere, Snowflake IDs, Redisson, Caffeine
• Frontend: Vue 3, Arco Design, Axios, QRCode.js, Pinia, OpenAPI; Taro+React for mini‑app