Re: [Boston.pm] function that takes two arguments and passes them into s///

Maybe try
cleanup('prefix_(\w+)', 'PREFIX_$1');
  as
    cleanup(  qr{prefix_(\w+)},   'PREFIX_$1'  );
?

You do NOT need the /e flag(s) for the $1 substitution to work.
I don't know if it works if it is in a substituted string or if it has 
to be top level of the parse.
The capturing parens need to be seen as RegEx special chars when parsed 
(as does the \w+),

(the alternative is to build the whole s/// statement in a string and 
eval _that_.)

it will have a better chance of working
On 01/30/2019 01:36 PM, Greg London wrote:
> Tried to boil this down to the smallest testcase i could come up with.
>
> I have a function that wraps a s/// substitution.
> Callers pass in the pattern and the replacement string.
> There's a bunch of other stuff going on that is why the
> function exists, but isn't why it isn't working.
>
> It's not working when I try pass in $1 as a substitution string.
> Cant get it to eval 'PREFIX_$1' inside the s/// piece.
>
> i've tried s///e and s///ee and neither work.
> I've tried qq()'ing the text, tryign to turn it into an expression, etc.
> Everything I've tried, I get one of two things in the output:
> Either the output has literally PREFIX_$1 in the result
> or it completely removes teh previx_nn from the result.
>
> I've used eval() on '$var' in other settings and it works.
> But cant seem to get it to work in the case of the replacement
> string in a s/// substitution.
>
> Thoughts?
>
> I'm using perl 5.8, if it matters.
>
> Here's the test code:
>
>
> #!/usr/bin/perl
>
> use warnings;
> use strict;
>
> #---------------------------------------
> # modules and other library code
> #---------------------------------------
>
> my $content =<<'CONTENT';
> some prefix string
> HERE: prefix_01
> post prefix stuff
> CONTENT
>
>
> sub cleanup{
> 	my($oldtext, $newtext)=@_;
> 	my $original=$content;
> 	my $modified=$content;
> 	my $changed_flag=0;
>
> 	if($modified =~ s/$oldtext/$newtext/ge  ){
> 		$changed_flag=1;
> 	}
>
> 	print "\n\n################################################\n";
> 	print "oldtext is '$oldtext'\nnewtext is '$newtext'\noriginal:
> \n$original\n";
> 	print "modified: \n$modified\nchange_flag is $changed_flag\n";
>
> }
>
> #---------------------------------------
> # user code
> #---------------------------------------
>
> # straightforward pattern with no capture: works
> cleanup('prefix','PREFIX');
>
> # using capture parens and $1: does not work
> cleanup('prefix_(\w+)', 'PREFIX_$1');
>
> _______________________________________________
> Boston-pm mailing list
> [email protected]
> https://mail.pm.org/mailman/listinfo/boston-pm
>

-- 
Bill Ricker
independent contractor for The Perl Shop


_______________________________________________
Boston-pm mailing list
[email protected]
https://mail.pm.org/mailman/listinfo/boston-pm