Tuple update (#2235)

* 'next_float' macro patch

More optimized implementation.

* tuple-type update

* U suffix in 'next_float'

* Do not add triplet, quadruplet, keep Tuple but deprecate. Add unpack

---------

Co-authored-by: Christoffer Lerno <christoffer@aegik.com>

lib/std/collections/tuple.c3

@@ -1,11 +1,24 @@
-module std::collections::tuple{Type1, Type2};
+module std::collections::pair{Type1, Type2};
 
-struct Tuple
+struct Pair
 {
 	Type1 first;
 	Type2 second;
 }
 
+<*
+ @param [&out] a
+ @param [&out] b
+ @require $assignable(self.first, $typeof(*a)) : "You cannot assign the first value to a"
+ @require $assignable(self.second, $typeof(*b)) : "You cannot assign the second value to b"
+*>
+
+macro void Pair.unpack(&self, a, b)
+{
+	*a = self.first;
+	*b = self.second;
+}
+
 module std::collections::triple{Type1, Type2, Type3};
 
 struct Triple
@@ -13,4 +26,12 @@ struct Triple
 	Type1 first;
 	Type2 second;
 	Type3 third;
-}
+}
+
+module std::collections::tuple{Type1, Type2};
+
+struct Tuple @deprecated("Use 'Pair' instead")
+{
+	Type1 first;
+	Type2 second;
+}

lib/std/math/random.c3

@@ -117,7 +117,7 @@ macro bool next_bool(random)
 *>
 macro float next_float(random)
 {
-	uint val = random.next_int() & (1 << 24 - 1);
+	uint val = random.next_int() & (1U << 24 - 1);
 	return val * 0x1.0p-24f;
 }