D, or "C++ Done Right(er)"
One of the languages that I spend a lot of time using is the D Programming Language. I’ve gotten quite a bit of (mostly) good-natured ribbing at my choice of language, but I normally just respond by showing people some of the more interesting features that I consider to be a huge improvement over C++, such as:
- Templates that are actually easily usable
- Compile-Time Function Execution, to do heavy processing at compile time instead of at run time
- United Function Call Syntax, allowing developers to simulate extending classes using a natural syntax
As a small example, last night another D programmer was asking about a way to do something similar to Python’s setdefault for dictionaries.
After a short setup period, I produced the following code:
import std.stdio;
ref T setdefault(T, U)(ref T[U] aa, U key, lazy T defaultvalue)
{
if(key !in aa) aa[key] = defaultvalue();
return aa[key];
}
class A
{
this(int i)
{
writeln("A's constructor ", i);
}
}
void main()
{
A[string] aa;
aa.setdefault("t2", new A(3));
aa.setdefault("t2", new A(4));
auto b = aa.get("test", new A(2));
}
The key part is that my setdefault function could be used as if it were a member if the associative array - UFCS in action.
Some of my code examples and breakdowns will be in D, because it’s an interesting language that has some cool capabilities.