CODEWITHSUNDEEP
nestjs
plusheen
plusheen

Reputation: 1426

Why does nest.js use a factory to specify a type rather than the class itself?

It's common in Nest.JS to see something like this:

@SomeThing({ type: () => MyEntity })
class MyClass {}

What is the purpose of this? It can't be for preventing things from running, as its a class reference, not an instance, and the import is already run. Nor do I think it's lazy loading, as there's no lazy references to it.

So what's the reason for this pattern over, say:

@SomeThing({ type: MyEntity })
class MyClass {}

My objective is, I have a huge list of metadata that has this type factory as a property and I need the underlying entity it's pointing to. I want to know if there are downsides to calling the function to get the entity

e.g.

const targetClasses = someMetadata.map(meta => meta.targetFn()

Upvotes: 0

Views: 218

Answers (1)

Jay McDoniel
Jay McDoniel

Reputation: 70432

It actually is for lazy loading, but not necessarily for the reason you think. It's to ensure that there's no problems when it comes to circular references. If Nest just used type: MyEntity and MyEntity imported MyClass, then one of the two files (or possibly both of them) would have that type show up as undefined. By allowing the type to resolve lazily using a factory, it ensures that the type value is there. This also prevents overhead from remembering when to use type: Class vs type: () => Class by just unifying the API

Upvotes: 1

Related Questions