NAME

Syntax::Operator::In - infix element-of-list meta-operator

SYNOPSIS

On a suitably-patched perl:

use Syntax::Operator::In;

if($x in<eq> @some_strings) {
   say "x is one of the given strings";
}

DESCRIPTION

This module provides an infix meta-operator that implements a element-of-list test on either strings or numbers.

Current versions of perl do not directly support custom infix operators. The documentation of XS::Parse::Infix describes the situation, with reference to a branch experimenting with this new feature. This module is therefore almost entirely useless on standard perl builds. While the regular parser does not support custom infix operators, they are supported via XS::Parse::Infix and hence XS::Parse::Keyword, and so custom keywords which attempt to parse operator syntax may be able to use it.

For operators that already specialize on string or numerical equality, see instead Syntax::Operator::Elem.

OPERATORS

in

my $present = $lhs in<OP> @rhs;

Yields true if the value on the lefhand side is equal to any of the values in the list on the right, according to some equality test operator OP.

At present this test operator must be either eq for string match, or == for number match.

TODO

• Improve runtime performance of compiletime-constant sets of strings, by detecting when the RHS contains string constants and convert it into a hash lookup.

• Consider further on the syntax for this operator. Maybe instead of the circumfix anglebrackets, a single colon might look nicer?

  $lhs in:OP @rhs
  
  $x in:== @numbers
  $x in:eq @strings

  Does the lacking of end marker make parsing harder though?

• Use XS::Parse::Infix support for defining new infix operators and permit cross-module integration with Syntax::Operator::Equ:

  $x in<===> @numbers-or-undef    # or  in:===
  $x in<equ> @strings-or-undef    # or  in:equ

• Consider cross-module integration with Syntax::Keyword::Match, permitting

  match($val : elem) {
     case(@arr_of_strings) { ... }
  }

  Or perhaps this would be too weird, and maybe match/case should have an "any-of" list/array matching ability itself. See also https://rt.cpan.org/Ticket/Display.html?id=143482.

AUTHOR

Paul Evans <leonerd@leonerd.org.uk>