Which is the best pattern to use in the controller?
Background
The REST API layer is still under design. Different ideas has been proposed up to now, but still we are not sure which is the best one to adopt for the project.
Initally, the - OP-4Getting issue details... STATUS was proposed. Then the - OP-6Getting issue details... STATUS has started and divided into a number of subtasks.
One of these subtask (the - OP-118Getting issue details... STATUS ) changed the initial proposal to something different while not yet completed (premature merge) and while other contributors were relying on the initially proposed pattern in other isses (e.g. - OP-178Getting issue details... STATUS and - OP-127Getting issue details... STATUS )
So now the point is to adopt one of the two, with an agnostic comparison between different solutions.
Relevant data
almost 90% of ModelDTOs in “api” would be identical to Models in “core”
some ModelDTO (10%) may need a dedicated mapping
the Swagger documentation need to be coherent with the code
Options considered
Option 1: | Option 2: | Option 3: | |
---|---|---|---|
Description | Linear mapping | Automapping | AbstractController |
Pros and cons | doesn’t need more knowledge for Java Base contributors (more linear) makes the controllers ticker | offer a centralized point for models mapping can be customized or disabled where not required it needs annotations (less linear) it binds to a 1:1 mapping that may be useless in the future | it should simplify the writing of a standard controller it works only for controllers managing one Model/DTO it binds to a 1:1 mapping that may be useless in the future it may lead to a “GOD Class” known problem |
Estimated cost | SMALL | SMALL | SMALL |
Action items
These are the proposed steps in order to take a decision
- Riccardo Costa
continues with the linear mapping on module “OPD” as per his own OP-118 - Uni2grow Cameroun continues with his own work on the OP-115
- Antonio Verni propose a change to the initial pattern in order to make it not exlusive and applying it on other OP-6 unassigned subtasks (e.g. OP-119)
- Vito Romano continues his own work on OP-116 trying using the the change to the initial pattern coming from Antonio
- Other Contributors to feel free to follow one of the two patterns or to propose a different one in their developments
- Alessandro Domanico To decide the final pattern before the end of March