New pattern matching doc

[Remi Forax]

> De: "Brian Goetz" <brian.goetz at oracle.com>
> À: "amber-spec-experts" <amber-spec-experts at openjdk.java.net>
> Envoyé: Mercredi 6 Janvier 2021 21:52:19
> Objet: New pattern matching doc

> I have pushed

> [
> https://github.com/openjdk/amber-docs/blob/master/site/design-notes/pattern-match-object-model.md
> |
> https://github.com/openjdk/amber-docs/blob/master/site/design-notes/pattern-match-object-model.md
> ]

> to the amber-docs repo, which outlines the direction for how patterns fit into
> the object model.
Hi Brian, 
i'm glag that "deconstructor" can now be virtual. 

Given that a "deconstructor" is now very like a method but with the caveat that you can not call them directly, 
i think it's time to change the name from "deconstructor" to xxx method, let say "case method" (i'm sure there is a better name), because they are closer to real methods than initially anticipated. 

Otherwise, "adhoc polymorphism" is a kind of overloaded ( :) ) term, it is used for overloading, and several other exotic kinds of polymorphism in the literature, 
i think that pattern polymorphism or case polymorphism is a better term. 

I think there is a missing discussion about what a case method can return, it can return something saying no match, it can return something saying there is a match and here are the values. 
One question is how the values are associated to the bindings, in the pattern, you have a notion of order, i.e. Point(var x, var y), x is the first binding and y the second, but in your proposed syntax, there is no notion of order, you associate a name to a value. 
If you remember, that's why i ask if when a pattern calls the case method, it's a call by name or a call by value. 
Given that you want overloading, it has to be a call by value, but in that case, the proposed syntax looks very weird. 
I would prefer a syntax that convey that the semantics is not a call by name 

By example, something along that line 
class Optional<T> { 
T value; 

case of() { 
if (value == null) { 
return no-match; 
} else { 
return match(value); 
} 
} 

case empty() { 
if (value == null) { 
return match; 
} else { 
return no-match; 
} 
} 
} 

Here the "return" is redundant, but my point is not to discuss the syntax but the associated semantics, so the syntax is just here to illustrate the point. 

regards, 
Rémi