Weak Flag
Weak is a conceptual flag of design intent. Models have relationships, but sometimes they have a circular relationship. While OneGen doesn't mind that, the generated models might, or rather the programming language or database.
Imagine you're designing a Notes app. There are two models: User
and Note
.
User can have many notes, so we add 1:m relationship to the User: notes: [Note]
. Notes on
the other hand always belong to a User, so we'd add a 1:1 relationship to the Note: user: User
.
Understanding Strong and Weak
If you're not familiar with the concept of a strong/weak reference, we feel it'd be best to read about it for the programming language of your expertise. We'll do our best to explain the concept, but please keep in mind it won't correctly apply to all programming languages out there.
To simplify it, you could think of a strong reference being required while a weak reference is optional.
- By using a strong reference you're saying: I need to use this reference, so you better keep it in the memory for as long as I need it.
- By using a weak reference you're saying: I'd like to use this reference, but don't sweat it, if you need to clean up the memory, or if the referenced type is not active anymore, feel free to erase it. I'm not counting on it.
Analysing the issue
OneGen supports a variety of programming languages/technologies and plans to support a lot more. Even though some languages can deal with circular references on their own, we can't make a flat statement. OneGen cannot automatically flag one reference as weak because that may not align with your existing system requirements.
If your design contains circular relationships/references, only you can get a sense of the data flow and we
recommend always flagging one end as weak
. For engines where circular
references don't pose an issue, the generated code should be the same.