helper/schema: remove outdated ValidateFunc restrictions

[vancluever]

Validation on non-primitives has possibly worked for several years -
see initial commit in 2a179d1, and subsequent blames associated with
respective lines in functions validateType, validateList, and
validatePrimitive.

For some reason though, we've never removed the restriction from
internalValidate.

This commit removes it, in addition to fixing the specific test case,
which was giving a false positive on the validation error as it was
missing the Elem attribute for TypeSet.

[radeksimko]

👋
I'm a little bit worried about "opening the flood gates" without more thorough testing. I'm sure this works well for simple cases like set/list of strings (as demonstrated in the modified test), or other primitive types, but I'm not sure we have the right SDK API in place to provide any reasonable UX for validating deeply nested structures, i.e. cases where Elem is not a primitive type.

Technically it should just work, but I'd rather not expand the surface of the SDK for now, especially after the 0.12 work, which has left parts of this package practically untested.

For these reasons I'm tempted to close the PR, but I'm sure it's something we'd want to address eventually as part of our future work on the SDK, when we get to a point of better test coverage and feel safer to introduce such changes.

[vancluever]

Hey Radek,

This is a fair concern.

Maybe we can compromise and allow it on primitive element types but not nested ones?

PS: It should be noted that the only thing preventing someone from using ValidateFuncs in this manner is InternalValidate, which is only called when using the acceptance test framework. It actually doesn't prevent the use in practice - you can sidestep it by just building the binary and not using the acceptance tests.

[vancluever]

@radeksimko if you could weigh in on hashicorp/terraform-provider-tfe#86 (comment) that'd be great, as I'm kind of confused how what @lafentres is doing is working if we are still restricting this in InternalValidate?

[vancluever]

I'm going to close this. As per the aforementioned thread, this works just fine if you add the ValidateFunc to the individual element versus the actual TypeList/TypeSet.

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.