Great question! You’ve already nailed the core distinction between spring-boot-starter-web and spring-boot-starter-data-rest, but there are several more key differences and practical decision points that can help you pick the right fit for your project.

Additional Key Differences You Might Have Missed #

• Automatic API Generation vs. Full Custom Control
  spring-boot-starter-data-rest automatically builds complete CRUD REST endpoints for your JPA repositories—no controller code required. It introspects your entities and repos to create standard paths (like /products for a ProductRepository). Meanwhile, spring-boot-starter-web demands you manually write controllers, map request paths, and implement every piece of request handling logic. This gives you total control over API behavior, but means more boilerplate for basic CRUD tasks.

• Out-of-the-Box HATEOAS Support
  Data REST integrates Spring HATEOAS by default, so responses automatically include hyperlinks to related resources (aligning with REST maturity model Level 3). For example, a customer response might have links to their orders or a self-reference. With Web, you’ll need to explicitly configure HATEOAS beans and manually add links to your response DTOs if you want this hypermedia-driven behavior.

• Built-in Filtering, Sorting, and Pagination
  Data REST handles common query operations without extra code: filter results with parameters like ?category=electronics, sort with ?sort=price,desc, and paginate with ?page=2&size=15. For Web, you’ll need to add Pageable parameters to controller methods and build filtering logic yourself (or use tools like Spring Data Specifications).

• Granular Control Over Exposed Resources
  With Data REST, annotations like @RepositoryRestResource and @RestResource let you hide entire repositories, specific entity fields, or individual CRUD operations (e.g., disable the delete endpoint for a repo). For Web, you control exposed data entirely through your controller methods—no special annotations needed, just explicit choices about what to return.

• Extensibility Focus
  Web is built for complex business logic: you can easily integrate custom auth (using @PreAuthorize), handle file uploads, combine data from multiple sources, or implement non-CRUD operations (like /users/{id}/reset-password). Data REST’s extensibility centers on tweaking its auto-generated APIs—you can use RepositoryEventHandler for pre/post processing, but complex logic usually requires pairing it with a service layer or additional web controllers.

Decision Points to Guide Your Choice #

• Choose Data REST if:

  • You’re building a quick prototype or a simple CRUD-focused API and want to minimize boilerplate.
  • You want your API to follow HATEOAS standards without extra work.
  • Your use case is mostly standard CRUD with basic filtering/sorting needs.

• Choose Web if:

  • You need full control over API paths, request/response formats, or custom business logic (e.g., multi-step workflows, custom validation).
  • Your API includes non-CRUD operations or requires complex integration with other services.
  • You need fine-grained authorization control (e.g., different permissions for read vs. write on specific endpoints).
  • Your team prefers explicit controller-based patterns over auto-generated behavior.

内容的提问来源于stack exchange，提问作者shredding