Catch2, testing floats

In the previous Catch2 post we were talking about event listeners, today let me show how to test float numbers.

There is not much to say about float number testing in Catch2, it is straightforward. We can use Approx() object for that, this way

// testing float numbers
#define CATCH_CONFIG_MAIN
#include <catch.hpp>

TEST_CASE("we can use Approx to test floats", "[floats]") {
	double a = 1.0;
	double b = a + std::numeric_limits<double>::epsilon();
	REQUIRE_FALSE(b == a);
	REQUIRE(b == Approx(a));  // 1
	REQUIRE(b == Approx(a).epsilon(0.1));  // 2
}

Approx has three member functions epsilon(), margin() and scale() and bunch of operartor overloads. Function epsilon() can used to customize $\eps$ constant used during number comparsions, see 2 in the example above.