X++ operators

This article describes the operators supported in X++.

Assignment operators

An assignment changes the value of a variable or field. The following table shows the X++ assignment operators. There is no difference between prefix and postfix operators.

Code examples for assignment operators

// An example of assignment operators and their output. 
static void Example1()
{
    int i = 1;
    // Using the = operator. i is assigned the value of i, plus 1. i = 2.
    i = i + 1;
    info(strFmt("Example 1: The result is "), i); // The result is 2.
}

static void Example2()
{
    int i = 1;
    // Using the += operator. i is assigned the value of i, plus 1. 
    // i = 2 (i = i + 1).
    i += 1;
    info(strFmt("Example 2: The result is "), i); // The result is 2. 
}

static void Example3()
{
    int i = 1;
    // Using the ++ operator. i is incremented by 1, and then 
    // by 1 again in the second statement. The final value of i is 3.
    i++;
    ++i;
    info(strFmt("Example 3: The result is "), i); // The result is 3. 
}

static void Example4()
{
    int i = 1;
    // Using the -= operator. i is assigned the value of i minus 1. 
    // i = 0 (i = i - 1).
    i -= 1;
    info(strFmt("Example 4: The result is "), i); // The result is 0. 
}

static void Example5()
{
    int i = 1;
    // Using the -- operator. i is decremented by 1, and then by 
    // 1 again in the second statement. The final value of i is -1.
    i--;
    --i;
    info(strFmt("Example 5: The result is "), i); // The result is -1. 
}

Arithmetic operators

You use arithmetic operators to perform numeric calculations. Most of the operators are binary and take two operands. However, the not (~) operator is unary and takes only one operand. Syntax for binary operators: expression1 ArithmeticOperator expression2 Syntax for unary operators: ArithmeticOperator expression1

Code examples for arithmetic operators

int a = 1 << 4;      // Perform four left shifts on 1 (1*2*2*2*2). a=16.
int b = 16 >> 4;     // Perform four right shifts on 16 (16/2/2/2/2). b=1.
int c = 4 * 5;       // Multiply 4 by 5. c=20.
int d = 20 / 5;      // Divide 20 by 5. d=4.
int e = 100 div 21;  // Return the integer division of 100 by 21. e=4 (4*21 = 84, remainder 16).
int f = 100 mod 21;  // Return the remainder of the integer division of 100 by 21. f=16.
int g = ~1;          // Binary negate 1 (all bits are reversed). g=-2.
int h = 1 & 3;       // Binary AND. Return the bits that are in common in the two integers. h=1.
int i = 1 | 3;       // Binary OR. Return the bits that are set in either 1 or 3. i=3.
int j = 1 ^ 3;       // Binary XOR. Return the bits that are set in 1 and NOT set in 3, and vice versa. j=2.
int k = 1 + 3;       // Add 1 and 3. k=4.
int l = 3 - 1;       // Subtract 1 from 3. l=2.
int m = (400 > 4) ? 1 : 5;  // If 400>4, 1 is returned. Otherwise, 5 is returned. Because 400>4, 1 is returned. m=1.

Expression operators

The as and is expression operators control downcast assignments. Downcast assignments involve class or table inheritance. Assignment statements that implicitly downcast can cause errors that are difficult to predict and diagnose. You can use the as keyword to make your downcasts explicit. You can use the is keyword to test whether a downcast is valid at run time.

The as keyword

Use the as keyword for assignments that downcast from a base class variable to a derived class variable. The as keyword tells other programmers and the compiler that you believe that the downcast will be valid during run time.

• The compiler reports an error for downcast assignment statements that lack the as keyword.
• At run time, the as keyword causes the downcast assignment statement to assign null if the downcast isn't valid.
• This is keyword is often used to safely test whether the as keyword will work.

Code example for the as keyword

In the following code example, the DerivedClass class extends the BaseClass class. The code example contains two valid assignments between its basec and derivedc variables. The upcast assignment to basec doesn't require the as keyword, but the downcast assignment to derivedc does require the as keyword. The following code will compile and run without errors.

static void AsKeywordExample()
{
    // DerivedClass extends BaseClass.
    BaseClass basec;
    DerivedClass derivedc;
    // BottomClass extends DerivedClass.
    BottomClass bottomc;
    derivedc = new DerivedClass();
    // AS is not required for an upcast assignment like this.
    basec = derivedc;
    // AS is required for a downcast assignment like this.
    derivedc = basec as DerivedClass;
    bottomc = new BottomClass();
    // AS causes this invalid downcast to assign null.
    bottomc = basec as DerivedClass;
}

The is keyword

The is keyword verifies whether an object is a subtype of a specified class. The is expression returns true if the object is a subtype of the class, or if the object is the same type as the class. The compiler reports an error if an is keyword expression compares two types, but neither type is a subtype of the other, and they aren't of the same type. The compiler reports a similar error for any plain assignment statement between two types, where neither type is a subtype of the other, and they aren't of the same type. At run time, the type of variable that references the underlying object is irrelevant to the is keyword. The is keyword causes the system to verify the object that the variable references, not the declared type of the variable that references the object.

Code examples for the is keyword

The following code examples illustrate the conditions that control whether an is expression returns true or false. The code examples depend on the fact that the Form class and the Query class both extend the TreeNode class.

// The compiler issues an error for the following code. 
// The compiler ascertains that the Form class and the Query class are not 
// part of the same inheritance hierarchy. Both the Form class and the Query class
// extend the TreeNode class, but neither Form nor Query is a subtype of the other.
Form myForm = new Form();
info(strFmt("%1", (myForm is Query)));

// The Infolog displays 0 during run time, where 0 means false. No supertype 
// object can be considered to also be of its subtype class.
TreeNode myTreeNode = new TreeNode();
info(strFmt("%1", (myTreeNode is Form)));

// The Infolog displays 0 during run time, where 0 means false. A null 
// reference causes the is expression to return false.
Form myForm;
info(strFmt("%1", (myForm is Form)));

// The Infolog displays 1 during run time, where 1 means true. 
// An object is an instance of its own class type.
Form myForm = new Form();
info(strFmt("%1", (myForm is Form)));

// The Infolog displays 1 during run time, where 1 means true. 
// Every subtype is also of its supertype.
Form myForm = new Form();
info(strFmt("%1", (myForm is TreeNode)));

// The Infolog displays 1 during run time, where 1 means true. 
// The type of the underlying object matters in the is expression,
// not the type of the variable that references the object.
Form myForm = new Form();
TreeNode myTreeNode;
myTreeNode = myForm; // Upcast.
info(strFmt("%1", (myTreeNode is Form)));

Code example for the is and as keywords

The following code example contains a typical use of the is keyword. The as keyword is used after the is keyword verifies that the as keyword will succeed. In this example, the is and as keywords are uppercase to make them more visible.

static void IsKeywordExample() 
{
    DerivedClass derivedc;
    BaseClass basec;
    basec = new DerivedClass();  // An upcast.
    if (basec IS DerivedClass)
    {
        info("Test 1: (basec IS DerivedClass) is true. Good.");
        derivedc = basec AS DerivedClass;
    }
    basec = new BaseClass();
    if (!(basec IS DerivedClass))
    {
        info("Test 2: !(basec IS DerivedClass) is true. Good.");
    }
}

//Output to the Infolog
Test 1: (basec IS DerivedClass) is true. Good.
Test 2: (!(basec IS DerivedClass)) is true. Good.

Object class as a special case

The Object class can appear as a special case in inheritance functionality. The compiler bypasses type checking for assignments to and from variables that are declared to be of type Object. Some classes inherit from the Object class, some classes inherit from another class, and some classes don't inherit from any class. Although the Dialog class doesn't inherit from any class, the assignment and call statements in the following code example work. However, if the assignment is bank4 = dlog3;, it will fail at compile time, because the Bank and Dialog classes have no inheritance relationship to each other. The compiler performs only one small validation on assignments to a variable that is declared to be of the Object class. The compiler verifies that the item that is being assigned to the Object variable is an instance of a class. The compiler doesn't allow an instance of a table buffer to be assigned to the Object variable. Additionally, the compiler doesn't allow primitive data types, such as int or str, to be assigned to the Object variable.

static void ObjectExample()
{
    Bank bank4;
    Object obj2;
    Dialog dlog3 = new Dialog("Test 4.");
    obj2 = dlog3;  // The assignment does work.
    obj2.run(false);  // The call causes the dialog to appear.
    info("Test 4a is finished.");
}

Tables

All tables inherit directly from the Common system table, unless they explicitly inherit from a different table. The Common table can't be instantiated. It doesn't exist in the underlying physical database. The Common table inherits from the xRecord class, but in a special way that isn't appropriate for the is keyword or the as keyword. When the as keyword is used to perform an invalid downcast among tables, the target variable references an unusable non-null entity. Any attempt to de-reference the target variable will cause an error that stops the program.

The is and as keywords and extended data types

Each extended data type has an Extends property. The style of inheritance that this property controls differs from the style of inheritance that the is and as keywords are designed for.

Relational operators

The following table lists the relational operators that can be used in X++. Most of the operators are binary and take two operands. However, the not (!) operator is unary and takes only one operand. Syntax for binary operators: expression1 relationalOperator expression2 Syntax for unary operators: relationalOperator expression1

The like operator

The like operator can use * as a wildcard character for zero or more characters, and ? as a wildcard character for one character. The maximum length of the operand is 1,000 characters. The like operator is evaluated by the underlying SQL, so the result might differ on different installations. If the expressions that you're comparing contain a file path, you must include four backslashes between each element, as shown in the following example.

select * from xRefpaths
where xRefPaths.Path like "\\\\Classes\\\\AddressSelectForm"

The equal (==) operator

When you use the equal (==) operator to compare objects, the object references are compared, not the objects themselves. This behavior might cause issues if you compare two objects, one of which is located on the server, and the other of which is located on the client. In these cases, you should use the equal method in the Object class. You can override this method to specify what it means for two objects to be equal. If you don't override the equal method, the comparison is identical to the comparison that is done by the equal (==) operator.

Code examples for relational operators

"Jones" like "Jo?es"  // Returns true, because the ? is equal to any single character.
"Fabrikam, Inc." like "Fa*"  // Returns true, because the * is equal to zero or more characters.
(( 42 * 2) == 84)  // Returns true, because 42*2 is equal to 84.
today() >= 1\1\1980  // Returns true, because today is later than January 1, 1980.
((11 div 10) >= 1)  // Returns true, because 11 div 10 is 1 (therefore, >= 1 is true).
(11<= 12)  // Returns true, because 11 is less than 12.
((11 div 10) > 1)  // Returns false, because 11 div 10 is 1.
(11 div 10) < 1)  // Returns false, because 11 div 10 is 1.
(11 != 12)  // Returns true, because 11 is not equal to 12.
(1 == 1) && (3 > 1)  // Returns true, because both expressions are true.

Operator precedence

The order that a compound expression is evaluated in can be important. For example, (x + y / 100) gives a different result, depending on whether the addition or the division is done first. You can use parentheses (()) to explicitly tell the compiler how it should evaluate an expression. For example, you can specify (x + y) / 100. If you don't explicitly tell the compiler the order that you want operations to be done in, the order is based on the precedence that is assigned to the operators. For example, the division operator has higher precedence than the addition operator. Therefore, for the expression x + y / 100, the compiler evaluates y / 100 first. In other words, x + y / 100 is equivalent to x + (y / 100). To make your code easy to read and maintain, be explicit. Use parentheses to indicate which operators should be evaluated first. The following table lists the operators in order of precedence. The higher an operator appears in the table, the higher its precedence. Operators that have higher precedence are evaluated before operators that have lower precedence. Note that the operator precedence of X++ isn't the same as the operator precedence of other languages, such as C# and Java.

Operators on the same line have equal precedence. If an expression includes more than one of these operators, it's evaluated from left to right, unless assignment operators are used. (Assignment operators are evaluated from right to left.) For example, && (logical AND) and || (logical OR) have the same precedence, and are evaluated from left to right. Therefore:

• 0 && 0 || 1 is equal to 1
• 1 || 0 && 0 is equal to 0.