Bug 9551 – template this parameter not recognized in constructors

The template this parameter, which works fine for normal methods: class Base { string label; this(string label_) { label = label_; } string toString(this This_)() { return This_.stringof ~ "(`" ~ label ~ "`)"; } } class Derived: Base { this() { super("Hello, world!"); } } unittest { Derived d = new Derived(); assert(d.toString() == "Derived(`Hello, world!`)"); } doesn't work for constructors: class Base { this(this This_)() { // Build a dispatch table using __traits(allMethods, This_) } void opCall(Payload_)(Payload_ payload_) { // Dynamically dispatch payload_ using the previously built dispatch table. } } class Derived: Base { // implicitly generated constructor will call the Base.this() and implicitly give it the static type of Derived. } unittest { Base b = new Derived; b(); // 100% transparent dynamic dispatch } Because of the following compile-time errors: C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: found 'This_' when expecting ')' C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: semicolon expected following function declaration C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: Declaration expected, not ')'

(In reply to comment #0) > Because of the following compile-time errors: > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: found 'This_' when > expecting ')' > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: semicolon expected > following function declaration > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: Declaration expected, not > ')' As far as I see, it conflicts with postblit syntax `this(this)`.

(In reply to comment #1) > (In reply to comment #0) > > Because of the following compile-time errors: > > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: found 'This_' when > > expecting ')' > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: semicolon expected > > following function declaration > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: Declaration expected, not > > ')' > > As far as I see, it conflicts with postblit syntax `this(this)`. Yes, it definitely looks like it, but the existence of a type name after "this" in the template parameter list should disambiguate it.

(In reply to comment #2) > (In reply to comment #1) > > (In reply to comment #0) > > > Because of the following compile-time errors: > > > > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: found 'This_' when > > > expecting ')' > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: semicolon expected > > > following function declaration > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: Declaration expected, not > > > ')' > > > > As far as I see, it conflicts with postblit syntax `this(this)`. > > Yes, it definitely looks like it, but the existence of a type name after "this" > in the template parameter list should disambiguate it. But, it is not just only a parser problem. Constructor will be treated specially for object construction entry. And, unfortunately, language semantics between non-mutable object construction and qualified constructor is yet not defined well. The combination of TemplateThisParameter and constructor may touch similar *yet not defined semantics*. Therefore, I must say that this issue will not be fix soon.

(In reply to comment #3) > (In reply to comment #2) > > (In reply to comment #1) > > > (In reply to comment #0) > > > > Because of the following compile-time errors: > > > > > > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: found 'This_' when > > > > expecting ')' > > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: semicolon expected > > > > following function declaration > > > > C:\Users\g.gyolchanyan\Desktop\test.d(3): Error: Declaration expected, not > > > > ')' > > > > > > As far as I see, it conflicts with postblit syntax `this(this)`. > > > > Yes, it definitely looks like it, but the existence of a type name after "this" > > in the template parameter list should disambiguate it. > > But, it is not just only a parser problem. > Constructor will be treated specially for object construction entry. And, > unfortunately, language semantics between non-mutable object construction and > qualified constructor is yet not defined well. The combination of > TemplateThisParameter and constructor may touch similar *yet not defined > semantics*. > > Therefore, I must say that this issue will not be fix soon. This is extremely unfortunate, because this is the only way that I know of to implement transparent dynamic dispatch.

(In reply to comment #4) > > This is extremely unfortunate, because this is the only way that I know of to > implement transparent dynamic dispatch. Couldn't you use helper function for the construction?

(In reply to comment #5) > (In reply to comment #4) > > > > This is extremely unfortunate, because this is the only way that I know of to > > implement transparent dynamic dispatch. > > Couldn't you use helper function for the construction? Theoretically, I could, but it would require immense boilerplate when dealing with large class hierarchies, which is hard to enforce.

(In reply to comment #6) > (In reply to comment #5) > > (In reply to comment #4) > > > > > > This is extremely unfortunate, because this is the only way that I know of to > > > implement transparent dynamic dispatch. > > > > Couldn't you use helper function for the construction? > > Theoretically, I could, but it would require immense boilerplate when dealing > with large class hierarchies, which is hard to enforce. I cannot imagine the 'immense boilerplate', but sorry it would need for a while.

(In reply to comment #1) > As far as I see, it conflicts with postblit syntax `this(this)`. Does it? postblit does not have the extra parentheses which is the signature of a template method. I'm asking out of ignorance.

(In reply to comment #8) > (In reply to comment #1) > > > As far as I see, it conflicts with postblit syntax `this(this)`. > > Does it? postblit does not have the extra parentheses which is the signature > of a template method. I'm asking out of ignorance. I'm saying that _currently_ they are confused, and should be properly distinguished.

(In reply to Gor Gyolchanyan from comment #0) > Base b = new Derived; > b(); // 100% transparent dynamic dispatch The code compiles with dmd v2.077.1 (if you add a dummy argument above, to match the signature, e.g. b(4)).