Building the Right Environment to Support AI, Machine Learning and Deep Learning
Although you do not need to write overloaded operators every day, this feature certainly puts the latest version of VB.NET (whether you call it 2.0 or 8.0there is some confusion in this area) on par with even the most powerful object-oriented languages. This article demonstrates a step-by-step process for writing custom operators, including a brief explanation for neophytes. Of course, if you didn't like VB.NET because of its differences from VB6, you probably won't get too excited about the ability to overload operators in VB.NET 2.0.
What Operators Are and Why You Overload Them
Programming has a rich history based in mathematics. Mathematics is steeped in meaningful symbols, operators, and operands. Operators are symbols that perform a function and operands are the things on which operators operate.
Operators typically indicate how many operands they use. Usually, you'll encounter unary, binary, and ternary operators. Unary means one operand, binary means two, and ternary means three. "Not" is an example of a unary operator; "+" is an example of a binary operator; and "?:" is an example of a ternary operator. (I have never seen a quaternary operator.)
Operators are convenient because they are quick to write and they are familiar, especially in arithmetic operations. For example, "&" and "+" are Boolean and arithmetic operators, but you can also used them to perform string concatenation for string operands. Clearly, operators have had multipleor overloadedmeanings for some time. A natural evolution is this ability extending to us programmers.
Basic Guidelines for Overloading Operators
You should follow a few basic guidelines for overloading operators to either get past the compiler or avoid confusing people:
- Don't change an operator's semantic meaning. The result of an operator should be intuitive.
- Make certain overloaded operators are shared methods.
- Overload operators in pairs, if a symmetric pair exists. These are the pairs of operators: "=" and "<>"; ">" and ">"; ">=" and "<="; and IsTrue and IsFalse.
- You cannot overload assignment (for example, the "=" operator in the preceding guideline is the equality-test operator.).
- You can overload only the operators listed in Table 1.
Table 1: Operators That Can Be Overloaded in .NET 2.0
Operator Operand Count Description + Unary Positive - Unary Negative IsFalse Unary Is False test IsTrue Unary Is True test Not Unary Negation + Binary Addition - Binary Subtraction * Binary Multiplication / Binary Floating-point division \ Binary Integer division & Binary Concatenation ^ Binary Exponentiation >> Binary Shift right << Binary Shift left = Binary Equality; assignment
cannot be overloaded
<> Binary Not equal > Binary Greater than < Binary Less than >= Binary Greater than or equal to <= Binary Less than or equal to And Binary Bitwise and Like Binary String pattern matching Mod Binary Modulo division Or Binary Bitwise or Xor Binary Bitwise xor CType Unary Type conversion
- Finally, don't assume that operators have been overloaded correctly or that others have defined all possible operators.
Defining a Named Method
Generic methods in VB.NET must be shared. They must have the same number of arguments as your operator will have. Thus, if you are overloading addition, you need a shared method with two arguments and a return type.
Suppose your problem domain needs a distance measured in feet and inches. (Don't worry; you could localize this class to switch to metric units, but that isn't relevant to this discussion.) To support your problem domain, you could easily define a class Distance that stores feet and inches as integers (see Listing 1).
Listing 1: A Distance Class That Stores Feet and Inches
Public Class Distance Private FFeet As Integer Private FInches As Integer Public Sub New(ByVal Feet As Integer, ByVal Inches As Integer) FFeet = Feet If (Inches > 12) Then FFeet += (Inches / 12) FInches = (Inches Mod 12) Else FInches = Inches End If End Sub Public Sub New(ByVal Inches As Integer) FFeet = (Inches / 12) FInches = (Inches Mod 12) End Sub Public Property Feet() As Integer Get Return FFeet End Get Set(ByVal value As Integer) FFeet = value End Set End Property Public Property Inches() As Integer Get Return FInches End Get Set(ByVal value As Integer) If (value > 12) Then FFeet += (value / 12) FInches = (value Mod 12) Else FInches = value End If End Set End Property Public Overrides Function ToString() As String Return FFeet & "ft " & FInches & "in" End Function End Class
Listing 1 converts all inches greater or equal to 12 to feet. For example, if you set the inches field using the Inches property, 14 inches becomes 1 foot, 2 inches. Next, you can support basic arithmetic operations by defining addition and subtraction. Following your named methods guidelines, you get the named methods in Listing 2.
Listing 2: Operations Implemented as Named Methods Separate Algorithm from Syntax
Public Shared Function Add(ByVal lhs As Distance, _ ByVal rhs As Distance) As Distance Return New Distance(lhs.Feet + rhs.Feet, lhs.Inches + rhs.Inches) End Function Public Shared Function Subtract(ByVal lhs As Distance, _ ByVal rhs As Distance) As Distance Return New Distance(lhs.Feet - rhs.Feet, lhs.Inches - rhs.Inches) End Function
After Listing 2, you can perform operations such as Distance.Add(Distance1, Distance2). All that remains is implementing the overloaded operators.