Rest API - 客户端与服务器

Question

I'm building an application where I have 2 types of entities: Template entity & Runnable entity. In the UI when I open template page I have 2 button 'Save' & 'Run'. Save updates the entity. Run creates runnable entity and sends him to run executor. I can also open runnable entity in the UI. for runnable entity: Save creates a new template entity and run creates another runnable entity and sends him to run executor. I'm wondering how to expose the save/run APIs and if the client should care on the type of entity it's currently working on. which option is better?

1. Create PUT API for save entity. In the PUT logic I'll check the type of the entity and decide if I can update or should I create another entity out of it and return this to the client. in this solution the client is agnostic to the type of entity it saves/runs
2. The client will need to know the type of entity it's currently trying to update/run. if it's trying to save template entity it will need to call PUT of TemplateResource and the server will return an error if the entity is actually a runnable entity. If it tries to save a runnable entity it will need to call POST for RunnableResource

Answer 1

I dont understand why on 2

The server will return an error if the entity is actually a runnable entity.

If I understand your spec correctly from the template page - Save does a PUT /api/template - while Run does a POST /api/runnable.

So there is no cunfusion here.

Then on the Run page: - Save does a POST /api/template - while Run does a POST /api/runnable.

So no confusion here either.

It is quite straitforward IMO since the HTTP verbs bear the intent. It seems simpler than implementing some logic server side to discover the intent.