Classic Edition: the symbol ⍠ is not available in Classic Edition, and the Variant operator is instead represented by ⎕U2360. Note too that ⍠ and ⎕OPT are synonymous though only the latter is available in the Classic Edition.
The Variant operator ⍠ specifies the value of an option to be used by its left operand function f. An option is a named property of a function whose value in some way affects the operation of that function.
For example, the Search and Replace operators include options named IC and Mode which respectively determine whether or not case is ignored and in what manner the input document is processed.
One of the set of options may be designated as the Principal option whose value may be set using a short-cut form of syntax as described below. For example, the Principal option for the Search and Replace operators is IC.
For the operand function with right argument Y and optional left argument X, the right operand B specifies the values of one or more options that are applicable to that function. If B is empty, function f is called with default options. Otherwise, B may be a scalar, a 2-element vector, or a vector of 2-element vectors which specifies values for one or more options as follows:
Option names and their values must be appropriate for the left operand function, otherwise DOMAIN ERROR (error code 11) will be reported.
tn←'Dick'(⎕FCREATE⍠'Z' 1)0
The following illustrations and examples apply to functions derived from the Search and Replace operators.
The following expression sets the IC option to 1, the Mode option to 'D' and the EOL option to 'LF'.
⍠('Mode' 'D')('IC' 1)('EOL' 'LF')
The following expression sets just the EOL property to 'CR'.
⍠'EOL' 'CR'
The following expression sets just the Principal option (which for the Search and Replace operators is IC) to 1.
⍠ 1
The order in which options are specified is typically irrelevant but if the same option is specified more than once, the rightmost one dominates. The following expression sets the option IC to 1:
⍠('IC' 0) ('IC' 1)
The Variant operator generates a derived function f⍠B and may be assigned to a name. The derived function is effectively function f bound with the option values specified by B.
The derived function may itself be used as a left operand to Variant to produce a second derived function whose options are further modified by the second application of the operator. The following sets the same options as the first example above:
⍠'Mode' 'D'⍠'IC' 1⍠'EOL' 'LF'
When the same option is specified more than once in this way, the outermost (rightmost) one dominates. The following expression also sets the option IC to 1:
⍠'IC' 0⍠'IC' 1
The following derived function returns the location of the word 'variant' within its right argument using default values for all the options.
f1 ← 'variant' ⎕S 0 f1 'The variant Variant operator' 4
It may be modified to perform a case-insensitive search:
(f1 ⍠ 1) 'The variant Variant operator' 4 12
This modified function may be named:
f2 ← f1 ⍠ 1 f2 'The variant Variant operator' 4 12
The modified function may itself be modified, in this case to revert to a case sensitive search:
f3 ← f2 ⍠ 0 f3 'The variant Variant operator' 4
This is equivalent to:
(f1 ⍠ 1 ⍠ 0) 'The variant Variant operator' 4
The Variant operator may also be used in conjunction with .NET classes; it can used to cast an array into a specific .NET data type, and to specify which constructor should be used when creating a new instance of a .NET class which has overloaded constructors. For further information, see Advanced Techniques.