Add initial tests for big rationals

core/math/big/api.odin

@@ -160,6 +160,5 @@ destroy :: proc {
 		int_destroy :: proc(integers: ..^Int)
 	*/
 	int_destroy,
-}
+	internal_rat_destroy,
-
+}
-

tests/core/math/big/build.bat

@@ -11,4 +11,6 @@ echo Running core:math/big tests
 echo ---
 
 %PATH_TO_ODIN% build . %COMMON% -o:speed -out:%OUT_NAME%
-python3 test.py %TEST_ARGS%
+python3 test.py %TEST_ARGS%
+
+odin test test_core_math_big.odin -file

tests/core/math/big/test_core_math_big.odin

@@ -35,3 +35,51 @@ test_permutations_and_combinations :: proc(t: ^testing.T) {
 		testing.expect_value(t, error, nil)
 	}
 }
+
+Rational_Vectors :: struct {
+	numerator:    int,
+	denominator:  int,
+	expected_f64: f64,
+	expected_f32: f32,
+	expected_f16: f16,
+	exact_f64:    bool,
+	exact_f32:    bool,
+	exact_f16:    bool,
+}
+rational_vectors := []Rational_Vectors{
+	{-1, 1, -1.00, -1.00, -1.00, true, true, true},
+	{ 1, 4,  0.25,  0.25,  0.25, true, true, true},
+	{ 3, 4,  0.75,  0.75,  0.75, true, true, true},
+	{-3, 4, -0.75, -0.75, -0.75, true, true, true},
+}
+
+@(test)
+test_rational_to_float :: proc(t: ^testing.T) {
+	for vec in rational_vectors {
+		r: big.Rat
+		defer big.destroy(&r)
+		big.set(&r.a, vec.numerator)
+		big.set(&r.b, vec.denominator)
+
+		{
+			float, exact, err := big.rat_to_f64(&r)
+			testing.expect_value(t, float, vec.expected_f64)
+			testing.expect(t, exact == vec.exact_f64)
+			testing.expect(t, err   == nil)
+		}
+
+		{
+			float, exact, err := big.rat_to_f32(&r)
+			testing.expect_value(t, float, vec.expected_f32)
+			testing.expect(t, exact == vec.exact_f32)
+			testing.expect(t, err   == nil)
+		}
+
+		{
+			float, exact, err := big.rat_to_f16(&r)
+			testing.expect_value(t, float, vec.expected_f16)
+			testing.expect(t, exact == vec.exact_f16)
+			testing.expect(t, err   == nil)
+		}
+	}
+}