Question

I have the following code in Perl:

if (index ($retval, $_[2]) != -1) {
    @fs = split ($_[2], $_[1]);

$_[2] is the delimiter variable and $_[1] is the string that the delimiter may exist in. ($_[0] is used elsewhere) You may have guessed that this code is in a subroutine by those variable names.

Anyway, onto my question, when my delimiter is something innocuous like 'a' or ':' the code works like it should. However, when it is something that would get parsed by Perl regex, like a '\' character, then it does not work like it is supposed to. This makes sense because in the split function Perl would see something like:

split (/\/, $_[1]); 

which makes no sense to it at all because it would want this:

split (/\//, $_[1]);

So with all of that in mind my question, that I cannot answer, is this: “How do I make it so that any delimiter that I put into $_[2], or all the ASCII characters, gets treated as the character it is supposed to be and not interpreted as something else?”

Thanks in advance,

Robert

Answer 1

You can use quotemeta to escape $_[2] properly so it will work in the regex without getting mangled. This should do it:

my $quoted = quotemeta $_[2];
@fs = split( $quoted, $_[1] );

Alternatively, you can use \Q in your regex to escape it. See “Escape Sequences” in perlre.