Re: Perl replacement for choice.com

[Athanasius]

in reply to Perl replacement for choice.com

A comment on the regexen:

# regex to add a comma after each char
# i believe you can think of it starting
# at the second character since
# ?<= means look to the left of the current
# position
# the . means to look at one character
# the () are required to capture each character
$show_choices =~ s/(?<=.)/,/g;
$show_choices =~ s/,$//g;       # remove the last comma

Actually, there is no capturing being done in the first regex; the parentheses are there only for grouping, i.e. to show where the positive look-behind assertion ends. (Try adding a $1 to the replacement part, and you will get a series of error messages beginning Use of uninitialized value $1.)

The first regex could also be written as:

$show_choices =~ s/(.)/$1,/g;

which does capture each character.

The two regexen can be combined into one using a negative look-ahead assertion:

$show_choices =~ s/(.)(?!$)/$1,/g;

See “Look-Around Assertions” under Extended Patterns in perlre.

Hope that helps,

[onelander]

Thank you very much for the correction. It is much appreciated.

[onelander]

Code updated to reflect the suggested changes.

[jwkrahn]

You don't have to capture a character at all:

$show_choices =~ s/(?<=.)(?=.)/,/g;