edecimal: Adaptive-Precision Decimal Integer

Why

Binary integers are the right representation for arithmetic and the wrong one for inspection: converting a large binary value to decimal for display costs a division per digit. edecimal stores the value in base 10 to begin with, so printing is free and every intermediate is exactly the number a human would write down.

That makes it the natural type for teaching material, for decimal-exact bookkeeping, and for cross-checking binary arithmetic against an independent representation.

#include <universal/number/edecimal/edecimal.hpp>
using namespace sw::universal;

edecimal a, b;
a = 123456789;
b = 987654321;
std::cout << a * b << '\n';   // 121932631112635269

What

edecimal is a non-template class holding a sign and a vector of decimal digits that grows with the value. There is no overflow.

Property	Value
Base	10
Width	grows with the value
Overflow	none
Representation	sign + decimal digit vector

Arithmetic, comparison, parsing and printing are provided, along with sqrt and the number-theoretic helpers used by the educational examples.

einteger – the binary equivalent, faster to compute with
dfloat – fixed-size IEEE 754-2008 decimal floating-point
dfixpnt – fixed-size decimal fixed-point