QuantityValue: Preserve precision of decimal values

Store decimal values internally and use that when explicitly casting to decimal, to preserve its precision
when constructing quantities such as Power, Information and BitRate.

Author: angularsen

UnitsNet/QuantityValue.cs

@@ -1,16 +1,16 @@
-// Copyright © 2007 Andreas Gullberg Larsen (angularsen@gmail.com).
+// Copyright (c) 2013 Andreas Gullberg Larsen (andreas.larsen84@gmail.com).
 // https://github.com/angularsen/UnitsNet
-// 
+//
 // Permission is hereby granted, free of charge, to any person obtaining a copy
 // of this software and associated documentation files (the "Software"), to deal
 // in the Software without restriction, including without limitation the rights
 // to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 // copies of the Software, and to permit persons to whom the Software is
 // furnished to do so, subject to the following conditions:
-// 
+//
 // The above copyright notice and this permission notice shall be included in
 // all copies or substantial portions of the Software.
-// 
+//
 // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 // IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 // FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
@@ -20,47 +20,87 @@
 // THE SOFTWARE.
 
 // Operator overloads not supported in Windows Runtime Components, we use 'double' type instead
+
 #if !WINDOWS_UWP
 using System;
 
 namespace UnitsNet
 {
     /// <summary>
-    ///     Pass it any numeric value (int, double, decimal, float..) and it will be implicitly converted to a <see cref="double" />, the quantity value representation used in UnitsNet.
-    ///     This is used to avoid an explosion of overloads for methods taking N numeric types for all our 500+ units.
+    ///     A type that supports implicit cast from all .NET numeric types, in order to avoid a large number of overloads
+    ///     and binary size for all From(value, unit) factory methods, for each of the 700+ units in the library.
+    ///     <see cref="QuantityValue"/> stores the value internally with the proper type to preserve the range or precision of the original value:
+    /// <list type="bullet">
+    /// <item><description><see cref="double"/> for [byte, short, int, long, float, double]</description></item>
+    /// <item><description><see cref="decimal"/> for [decimal] to preserve the 128-bit precision</description></item>
+    /// </list>
     /// </summary>
     /// <remarks>
-    ///     At the time of this writing, this reduces the number of From() overloads to 1/4th:
+    ///     At the time of this writing, this reduces the number of From(value, unit) overloads to 1/4th:
     ///     From 8 (int, long, double, decimal + each nullable) down to 2 (QuantityValue and QuantityValue?).
     ///     This also adds more numeric types with no extra overhead, such as float, short and byte.
     /// </remarks>
     public struct QuantityValue
     {
-        private readonly double _value;
+        /// <summary>
+        ///     Value assigned when implicitly casting from all numeric types except <see cref="decimal" />, since
+        ///     <see cref="double" /> has the greatest range and is 64 bits versus 128 bits for <see cref="decimal"/>.
+        /// </summary>
+        private readonly double? _value;
+
+        /// <summary>
+        ///     Value assigned when implicitly casting from <see cref="decimal" /> type, since it has a greater precision than
+        ///     <see cref="double"/> and we want to preserve that when constructing quantities that use <see cref="decimal"/>
+        ///     as their value type.
+        /// </summary>
+        private readonly decimal? _valueDecimal;
 
-        // Obsolete is used to communicate how they should use this type, instead of making the constructor private and have them figure it out
-        [Obsolete("Do not use this constructor. Instead pass any numeric value such as int, long, float, double, decimal, short or byte directly and it will be implicitly casted to double.")]
         private QuantityValue(double val)
         {
             _value = val;
+            _valueDecimal = null;
+        }
+
+        private QuantityValue(decimal val)
+        {
+            _valueDecimal = val;
+            _value = null;
         }
 
         #region To QuantityValue
 
 #pragma warning disable 618
-        public static implicit operator QuantityValue(double val) => new QuantityValue(val);
-        public static implicit operator QuantityValue(float val) => new QuantityValue(val);
-        public static implicit operator QuantityValue(long val) => new QuantityValue(val);
-        public static implicit operator QuantityValue(decimal val) => new QuantityValue(Convert.ToDouble(val));
-        public static implicit operator QuantityValue(short val) => new QuantityValue(val);
-        public static implicit operator QuantityValue(byte val) => new QuantityValue(val);
+        // Prefer double for integer types, since most quantities use that type as of now and
+        // that avoids unnecessary casts back and forth.
+        // If we later change to use decimal more, we should revisit this.
+        public static implicit operator QuantityValue(byte val) => new QuantityValue((double) val);
+        public static implicit operator QuantityValue(short val) => new QuantityValue((double) val);
+        public static implicit operator QuantityValue(int val) => new QuantityValue((double) val);
+        public static implicit operator QuantityValue(long val) => new QuantityValue((double) val);
+        public static implicit operator QuantityValue(float val) => new QuantityValue(val); // double
+        public static implicit operator QuantityValue(double val) => new QuantityValue(val); // double
+        public static implicit operator QuantityValue(decimal val) => new QuantityValue(val); // decimal
 #pragma warning restore 618
-        
+
         #endregion
 
         #region To double
 
-        public static explicit operator double(QuantityValue number) => Convert.ToDouble(number._value);
+        public static explicit operator double(QuantityValue number)
+        {
+            // double -> decimal -> zero (since we can't implement the default struct ctor)
+            return number._value.GetValueOrDefault((double) number._valueDecimal.GetValueOrDefault());
+        }
+
+        #endregion
+
+        #region To decimal
+
+        public static explicit operator decimal(QuantityValue number)
+        {
+            // decimal -> double -> zero (since we can't implement the default struct ctor)
+            return number._valueDecimal.GetValueOrDefault((decimal) number._value.GetValueOrDefault());
+        }
 
         #endregion
     }