LibCrypto: Add a constructor to (Un)SignedBigInteger taking a double

For now this will assume that the double given is exactly representable
as an integer, so no NaN, infinity or rounding.

Userland/Libraries/LibCrypto/BigInt/SignedBigInteger.h

@@ -43,6 +43,8 @@ public:
     {
     }
 
+    explicit SignedBigInteger(double value);
+
     [[nodiscard]] static SignedBigInteger create_invalid()
     {
         return { UnsignedBigInteger::create_invalid(), false };