Hi I was curious if there's an easy way to retrieve nested generics from a given type. For example say I have:
public class Token<T, V>
{
protected readonly T fieldT;
protected readonly V fieldV;
public Token(T arg1, V arg2)
{
fieldT = arg1;
fieldV = arg2;
}
}
public abstract class TokenBuilder<TokenType, T, V>
where TokenType: Token<T, V>
{
public abstract TokenType CreateToken(T arg1, V arg2);
}
This works well enough with subclassing. However when T, V are complex I'd prefer to infer them. For example
// T is a tuple (int num1, int num2, int num3, int num4)
// V is a tuple (string name1, string name2, string name3)
public class FancyToken : Token<
(int num1, int num2, int num3, int num4),
(string name1, string name2, string name3)
>
{
protected readonly (int num1, int num2, int num3, int num4) fieldT;
protected readonly (string name1, string name2, string name3) fieldV;
public Token(
(int num1, int num2, int num3, int num4) arg1,
(string name1, string name2, string name3) arg2
)
{
fieldT = arg1;
fieldV = arg2;
}
}
// All of this is fine!
// Do I have to redeclare <T, V> here? Shouldn't they be inferred from FancyToken?
public class FancyTokenBuilder : TokenBuilder<FancyToken, (int num1, int num2, int num3, int num4), (string name1, string name2, string name3)>
{
public FancyToken CreateToken((int num1, int num2, int num3, int num4) arg1, (string name1, string name2, string name3) arg2){
...
}
}
This is fairly minor, but I'm basically curious if the generic arguments supplied to FancyToken (T, V) can be "pulled out" and used by another class that takes in a Token as a parameter. Redeclaring these T, V explicitly works perfectly, but I'm curious if there's any way around re-writing those types again
