Question

When I use Boost.Tuple, I have to use some syntax like:

result.get<0>()

It looks very unfamiliar to me. Usually <> contains a typename, why does it use an int here?

Answer 1

The reason they use the syntax result.get<0>() is that each element of the tuple can have a different type, and this syntax is the simplest way in C++ to let the compiler do the right thing with types.

If the function were just plain get(0), all elements of the tuple would have to have the same type, because there’s no way to have one untemplated get function that returns several different types.

Something like result.get<int>(0) could theoretically work, but it’s more verbose, and introduces a potential source of error–what if the 0th element wasn’t an int at all? Worse still, you’d only be able to catch this error at runtime. The syntax used in Boost is plain and simple–the only way you can possibly screw it up is to specify an invalid index, and that can be caught at compile time.