Add support for custom errors

[fvictorio]

Closes #485. Way overdue 😅

[fvictorio]

The failing tests make no sense to me. I know I can fix them by running jest -u, but if we need to make that every time we add a node, maybe these tests aren't adding any value?

Plus, if something is really missing in those files, the "real" tests will fail anyway.

@Janther @mattiaerre do you mind if I remove those tests?

[fvictorio]

Ok, I preemptively removed them. I also excluded the scripts directory from the coverage check (which makes sense to me). But I'm totally fine with reverting that commit if you think it's a good idea to keep those tests.

[Janther]

This should be named FunctionCustomError

[Janther]

The problem with snapshots in testing is that the snapshots need to be understandable and auditable. A extremely big list or a massive array are not something that snapshots were made for.
I agree that we need to test the scripts but would make much more sense to create a sample folder with just a few files and make the test generate the snapshot for that folder. That would be more auditable.

[fvictorio]

The problem with snapshots in testing is that the snapshots need to be understandable and auditable. A extremely big list or a massive array are not something that snapshots were made for.
I agree that we need to test the scripts but would make much more sense to create a sample folder with just a few files and make the test generate the snapshot for that folder. That would be more auditable.

Do you mind if we do/discuss that in another PR so we can get this merged?

(FWIW, I'm OK with not testing code-generating scripts if the generated code is tested, and we already have a 100% threshold in the directory where the generated code is added)

[fvictorio]

Feedback applied, thank you @Janther!