ExplicitConstructors_05

The this() constructor initializer calls in this test and the following resulted in invalid code previously.

Is reusing the warning number okay here? #Resolved

Maybe, but reusing seems unnecessary. Updated.

hasAnyRecordConstructors

I think the name is confusing. The body is checking for a primary constructor, not just any constructor. #Closed

A 'this' initializer for a 'record struct' constructor must call the primary constructor or an explicitly declared constructor.

The wording looks too general. The requirement is only for records with primary constructors, I think the message should reflect that. See, for example, how we do that for UnexpectedOrMissingConstructorInitializerInRecord:

  <data name="ERR_UnexpectedOrMissingConstructorInitializerInRecord" xml:space="preserve">
    <value>A constructor declared in a record with parameter list must have 'this' constructor initializer.</value>
  </data>
``` #Closed

skipInitializer

Should we also check if !constructorSymbol.IsStatic? #Closed

public S4(object o) : this() { }

Consider also testing with static constructors. For example:

record struct S(int x)
{
    static S() : this()
    {}

    static int F = 3;
}
``` #Closed

skipInitializer

The fact that skipInitializer depends on includesFieldInitializers looks suspicious. I think it is technically possible to have a primary constructor and no initializers. It would be good trying to construct a scenario like that. Consider instead using condition thisInitializer && ContainingType.IsDefaultValueTypeConstructor(initializer), I think that would exactly reflect the condition that design requires us to check.
#Closed