Initialize decimal with literal

Question

I've got Decimal class which implements decimal floating point arithmetic. It can be initialized with an integral arithmetic type or a fractional digit stored as a string. Build-in floating point types are deliberately avoided because of their approximative nature. Such approach works well but fractional digits coded as strings look awkward. Is it possible to use somehow C++ literals to build strings (or binary representation) behind the scene?

Decimal a{"1.254684987"}; // current initialization
Decimal b{1.254684987_dec}; // desired way

score 5 · Answer 1 · edited Nov 16 '17 at 15:29

5

Given that you already have a c'tor taking a C-string, simply writr something like

Decimal operator "" _dec(const char* c){
    return Decimal{c}; // Assuming an explicit c'tor
}

auto whatever = 12.3_dec;

Of course, you need a C++11 compiler for this to work.

edited Nov 16 '17 at 15:29

iBug

35,554
7
89
134

answered Nov 16 '17 at 14:34

Massimiliano Janes

5,524
1
10
22

Wouldn't that work on C++11? – Daniel H Nov 16 '17 at 14:46
@DanielH yep, fixed – Massimiliano Janes Nov 16 '17 at 14:48
What's the name for this syntax? – iBug Nov 16 '17 at 14:51
@ibug in standardese, *raw literal operator* – Massimiliano Janes Nov 16 '17 at 14:54
@iBug More details are available [here](http://en.cppreference.com/w/cpp/language/user_literal "User-defined literals at cppreference"). – Daniel H Nov 16 '17 at 15:02
@DanielH I have that site in my bookmark. Anyway thank you. – iBug Nov 16 '17 at 15:16
2

oh... the case where you imagine an imaginary syntax that would just work for you and then you find out it isn't imaginary after all. If only there would be more of those moments... – bolov Nov 16 '17 at 15:33

Initialize decimal with literal

1 Answers1