4.3 KiB
4.3 KiB
brainstorm: start coding
Goal
Build the initial Java Web project scaffold and a login/permission skeleton for this JSP + Servlet + MySQL library-management system.
What I already know
- The user asked to start writing code.
- There is no active task before this one.
- The repository currently has Trellis project guidance but no application source code files.
- Existing project specs describe a B/S Java web application using JSP, Servlet, MySQL, Tomcat, and IDEA.
- Backend guidance expects a layered structure: JSP/CSS presentation -> Servlet controller -> Service/business -> DAO -> MySQL.
- Frontend guidance expects JSP/CSS pages, JSP includes or fragments, static assets, and small page scripts when needed.
- The user selected option A: build the Java Web foundation and login/permission skeleton first.
Assumptions (temporary)
- The desired work is to start building the library-management system rather than editing an existing application.
- Maven is a reasonable default for Java web project scaffolding unless the user prefers plain IDEA/Tomcat project files.
- The first implementation should establish a compilable structure and a minimal login flow rather than complete every business module.
Open Questions
- None currently blocking.
Requirements (evolving)
- Follow the existing Trellis backend and frontend specs.
- Keep backend and frontend conventions aligned with JSP + Servlet + MySQL, not SPA frameworks.
- Create or update tests/checks where practical for the selected implementation scope.
- Create the base Java Web project structure for Tomcat deployment.
- Add a login/permission skeleton covering administrator, librarian, and reader roles.
- Add model/entity, DAO, service, servlet/controller, JSP, and static asset structure needed for the login slice.
- Include database initialization or schema notes for user/role data needed by the skeleton.
- Keep business logic layered: Servlet parameter/session handling, Service authentication and role checks, DAO persistence access.
- Include safe error handling for invalid credentials, missing parameters, and unauthorized access.
Acceptance Criteria (evolving)
- The selected first module or feature is explicitly confirmed.
- The implementation follows the documented JSP + Servlet + MySQL layered architecture.
- A fresh checkout has recognizable Java Web/Tomcat project structure and build configuration.
- Login page submits credentials to a Servlet controller and displays validation/authentication failures safely.
- Authentication logic is represented through service and DAO boundaries rather than embedded in JSP.
- Session state stores the authenticated user and role in a controlled way.
- Basic role/permission constants or helpers exist for administrator, librarian, and reader.
- SQL/schema guidance exists for the minimal account/role tables needed by login.
- Lint, type-check, compile, or equivalent project validation is run where available.
Definition of Done (team quality bar)
- Tests added/updated where appropriate.
- Lint / typecheck / compile / CI-equivalent checks are green where available.
- Docs/notes updated if behavior changes.
- Rollout/rollback considered if risky.
Out of Scope (explicit)
- Choosing React, Vue, TypeScript, SPA routing, or frontend hook/state-library conventions unless the user explicitly changes the stack.
- Implementing full book, reader, borrowing, statistics, backup, or logging modules beyond placeholders needed for navigation or role skeleton.
- Implementing production-grade password reset, remember-me, MFA, or external identity provider integration.
Technical Notes
- Relevant spec indexes:
.trellis/spec/backend/index.md,.trellis/spec/frontend/index.md. - Backend pre-development checklist includes directory structure, database, error handling, logging, and quality guidelines.
- Frontend pre-development checklist includes directory structure, JSP component guidelines, state management, type safety, and quality guidelines.
- Codebase retrieval on 2026-04-27 found no application source code and surfaced the project specs as the main implementation context.
- Spec indexes reference
.trellis/tasks/00-bootstrap-guidelines/research/project-requirements.md, but that file is absent in the current workspace, so context curation uses only present spec files.