Fix Linux compatibility: Replace Darwin-specific pow() with Standards.power()

- Remove Darwin/Glibc platform-conditional imports
- Use Foundation-free FloatingPoint.power() method from Standards package
- Ensures cross-platform compatibility (macOS, Linux, iOS, etc.)
- Fixes CI build failures on Ubuntu

Resolves: IEEE_754.Scaling Tests compilation errors on Linux

Author: coenttb

Tests/IEEE 754 Tests/IEEE_754.Scaling Tests.swift

@@ -4,7 +4,7 @@
 // Comprehensive tests for IEEE 754-2019 Section 5.3 Scaling operations
 
 import Testing
-import Darwin
+import Standards
 
 @testable import IEEE_754
 
@@ -136,7 +136,7 @@ struct DoubleSignificandTests {
         let value = 12.34
         let sig = IEEE_754.Scaling.significand(value)
         let exp = IEEE_754.Scaling.exponent(value)
-        let reconstructed = sig * pow(2.0, exp)
+        let reconstructed = sig * 2.0.power(Int(exp))
         #expect(reconstructed == value, "value should equal significand × 2^exponent")
     }
 
@@ -223,7 +223,7 @@ struct FloatSignificandTests {
         let value: Float = 12.34
         let sig = IEEE_754.Scaling.significand(value)
         let exp = IEEE_754.Scaling.exponent(value)
-        let reconstructed = sig * Float(pow(2.0, Double(exp)))
+        let reconstructed = sig * Float(2.0).power(Int(exp))
         #expect(reconstructed == value, "value should equal significand × 2^exponent")
     }
 }