Next Previous Contents

7. Operators

S-Lang supports a variety of operators that are grouped into three classes: assignment operators, binary operators, and unary operators.

An assignment operator is used to assign a value to a variable. They will be discussed more fully in the context of the assignment statement in the section on Assignment Statements.

An unary operator acts only upon a single quantity while a binary operation is an operation between two quantities. The boolean operator not is an example of an unary operator. Examples of binary operators include the usual arithmetic operators +, -, *, and /. The operator given by - can be either an unary operator (negation) or a binary operator (subtraction); the actual operation is determined from the context in which it is used.

Binary operators are used in algebraic forms, e.g., a + b. Unary operators fall into one of two classes: postfix-unary or prefix-unary. For example, in the expression -x, the minus sign is a prefix-unary operator.

All binary and unary operators may be defined for any supported data type. For example, the arithmetic plus operator has been extended to the String_Type data type to permit concatenation between strings. But just because it is possible to define the action of an operator upon a data type, it does not mean that all data types support all the binary and unary operators. For example, while String_Type supports the + operator, it does not admit the * operator.

7.1 Unary Operators

The unary operators operate only upon a single operand. They include: not, ~, -, @, &, as well as the increment and decrement operators ++ and --, respectively.

The boolean operator not acts only upon integers and produces 0 if its operand is non-zero, otherwise it produces 1.

The bit-level not operator ~ performs a similar function, except that it operates on the individual bits of its integer operand.

The arithmetic negation operator - is perhaps the most well-known unary operator. It simply reverses the sign of its operand.

The reference (&) and dereference (@) operators will be discussed in greater detail in the section on Referencing Variables. Similarly, the increment (++) and decrement (--) operators will be discussed in the context of the assignment operator.

7.2 Binary Operators

The binary operators may be grouped according to several classes: arithmetic operators, relational operators, boolean operators, and bitwise operators.

Arithmetic Operators

The arithmetic operators include +, -, *, and /, which perform addition, subtraction, multiplication, and division, respectively. In addition to these, S-Lang supports the mod operator, which divides two numbers and produces the remainder, as as well as the power operator ^.

The data type of the result produced by the use of one of these operators depends upon the data types of the binary participants. If they are both integers, the result will be an integer. However, if the operands are not of the same type, they will be converted to a common type before the operation is performed. For example, if one is a floating point type and the other is an integer, the integer will be converted to a float. In general, the promotion from one type to another is such that no information is lost, if possible. As an example, consider the expression 8/5 which indicates division of the integer 8 by the integer 5. The result will be the integer 1 and not the floating point value 1.6. However, 8/5.0 will produce 1.6 because 5.0 is a floating point number.

Relational Operators

The relational operators are >, >=, <, <=, ==, and !=. These perform the comparisons greater than, greater than or equal, less than, less than or equal, equal, and not equal, respectively. For most data types, the result of the comparison will be a boolean value; however, for arrays the result will be an array of boolean values. The section on arrays will explain this is greater detail.

Note: For S-Lang versions 2.1 and higher, relational expressions such as a<b<=c are defined in the mathematical sense, i.e.,

      ((a < b) and (b <= c))
Simarily, (a < b <= c < d) is the same as
      ((a < b) and (b <= c) and (c < d))
and so on. In previous versions of S-Lang, (a<b<c) meant (a<b)<c; however this interpretation was not very useful.

Boolean Operators

S-Lang supports four boolean binary operators: or, and, ||, and &&, which for most data types, return a boolean result. In particular, the or and || operators return a non-zero value (boolean TRUE) if either of their operands are non-zero, otherwise they produce zero (boolean FALSE). The and and && operators produce a non-zero value if and only if both their operands are non-zero, otherwise they produce zero.

Unlike the operators && and ||, the and and or operators do not perform the so-called boolean short-circuit evaluation. For example, consider the expression:

      (x != 0) and (1/x > 10)
Here, if x were to have a value of zero, a division by zero error would occur because even though x!=0 evaluates to zero, the and operator is not short-circuited and the 1/x expression would still be evaluated. This problem can be avoided using the short-circuiting && operator:
     (x != 0) && (1/x > 10)
Another difference between the short-circuiting (&&,||) and the non-short-circuiting operators (and,or) is that the short-circuiting forms work only with integer or boolean types. In contrast, if either of the operands of the and or or operators is an array then a corresponding array of boolean values will result. This is explained in more detail in the section on arrays.

Note: the short-circuiting operators && and || were first introduced in S-Lang 2.1; they are not available in older versions.

Bitwise Operators

The bitwise binary operators are currently defined for integer operands and are used for bit-level operations. Operators that fall in this class include &, |, shl, shr, and xor. The & operator performs a boolean AND operation between the corresponding bits of the operands. Similarly, the | operator performs the boolean OR operation on the bits. The bit-shifting operators shl and shr shift the bits of the first operand by the number given by the second operand to the left or right, respectively. Finally, the xor performs an EXCLUSIVE-OR operation.

These operators are commonly used to manipulate variables whose individual bits have distinct meanings. In particular, & is usually used to test bits, | can be used to set bits, and xor may be used to flip a bit.

As an example of using & to perform tests on bits, consider the following: The jed text editor stores some of the information about a buffer in a bitmapped integer variable. The value of this variable may be retrieved using the jed intrinsic function getbuf_info, which actually returns four quantities: the buffer flags, the name of the buffer, directory name, and file name. For the purposes of this section, only the buffer flags are of interest and can be retrieved via a function such as

      define get_buffer_flags ()
      {
         variable flags;
         (,,,flags) = getbuf_info ();
         return flags;
      }
The buffer flags object is a bitmapped quantity where the 0th bit indicates whether or not the buffer has been modified, the first bit indicates whether or not autosave has been enabled for the buffer, and so on. Consider for the moment the task of determining if the buffer has been modified. This can be determined by looking at the zeroth bit: if it is 0 the buffer has not been modified, otherwise it has been modified. Thus we can create the function,
     define is_buffer_modified ()
     {
        variable flags = get_buffer_flags ();
        return (flags & 1);
     }
where the integer 1 has been used since it is represented as an object with all bits unset, except for the zeroth one, which is set. (At this point, it should also be apparent that bits are numbered from zero, thus an 8 bit integer consists of bits 0 to 7, where 0 is the least significant bit and 7 is the most significant one.) Similarly, we can create another function
     define is_autosave_on ()
     {
        variable flags = get_buffer_flags ();
        return (flags & 2);
     }
to determine whether or not autosave has been turned on for the buffer.

The shl operator may be used to form the integer with only the nth bit set. For example, 1 shl 6 produces an integer with all bits set to zero except the sixth bit, which is set to one. The following example exploits this fact:

     define test_nth_bit (flags, nth)
     {
        return flags & (1 shl nth);
     }

The Namespace Operator

The operator -> is used to in conjunction with a namespace to access an object within the namespace. For example, if A is the name of a namespace containing the variable v, then A->v refers to that variable. Namespaces are discussed more fully in the chapter on Namespaces.

Operator Precedence

Binary Operators and Functions Returning Multiple Values

Care must be exercised when using binary operators with an operand that returns multiple values. In fact, the current implementation of the S-Lang language will produce incorrect results if both operands of a binary expression return multiple values. At most, only one of operands of a binary expression can return multiple values, and that operand must be the first one, not the second. For example,

    define read_line (fp)
    {
       variable line, status;

       status = fgets (&line, fp);
       if (status == -1)
         return -1;
       return (line, status);
    }
defines a function, read_line that takes a single argument specifying a handle to an open file, and returns one or two values, depending upon the return value of fgets. Now consider
        while (read_line (fp) > 0)
          {
             text = ();
             % Do something with text
             .
             .
          }
Here the relational binary operator > forms a comparison between one of the return values (the one at the top of the stack) and 0. In accordance with the above rule, since read_line returns multiple values, it must occur as the left binary operand. Putting it on the right as in
        while (0 < read_line (fp))    % Incorrect
          {
             text = ();
             % Do something with text
             .
             .
          }
violates the rule and will result in the wrong answer. For this reason, one should avoid using a function that returns muliple return values as a binary operand.

7.3 Mixing Integer and Floating Point Arithmetic

If a binary operation (+, -, * , /) is performed on two integers, the result is an integer. If at least one of the operands is a floating point value, the other will be converted to a floating point value, and a floating point result be produced. For example:

      11 / 2           --> 5   (integer)
      11 / 2.0         --> 5.5 (double)
      11.0 / 2         --> 5.5 (double)
      11.0 / 2.0       --> 5.5 (double)

Sometimes to achive the desired result, it is necessary to explicitly convert from one data type to another. For example, suppose that a and b are integers, and that one wants to compute a/b using floating point arithmetic. In such a case, it is necessary to convert at least one of the operands to a floating point value using, e.g., the double function:

      x = a/double(b);

7.4 Short Circuit Boolean Evaluation

As of S-Lang version 2.1, use of the andelse and orelse have been deprecated in favor of the && and || short-circuiting operators.

The boolean operators or and and are not short circuited as they are in some languages. S-Lang uses orelse and andelse expressions for short circuit boolean evaluation. However, these are not binary operators. Expressions of the form:

expr-1 and expr-2 and ... expr-n
can be replaced by the short circuited version using andelse:
andelse {expr-1} {expr-2} ... {expr-n}
A similar syntax holds for the orelse operator. For example, consider the statement:
      if ((x != 0) and (1/x > 10)) do_something ();
Here, if x were to have a value of zero, a division by zero error would occur because even though x!=0 evaluates to zero, the and operator is not short circuited and the 1/x expression would be evaluated causing division by zero. For this case, the andelse expression could be used to avoid the problem:
      if (andelse
          {x != 0}
          {1 / x > 10})  do_something ();


Next Previous Contents