in reply to Issue with regex matching

you can escape the special chars stored in the variable $constant using \Q and \E. See also perlre.
if($_ =~ m/^(\Q$constant\E)/) { $line_count++; }
Since $constant is user specified you could simply call quotemeta on the user input before using it. Then you won't need \Q and \E in your regex. E.g.:
my $searchterm = quotemeta $user_input;
You are currently only counting lines which start with that pattern. You wrote that you wanted to count all occurrences of the pattern. If you want to match all the occurrences of $constant per line you could change your code to:
while ($_ =~ m/\Q$constant\E/gc) { $count++; }
Other thoughts:

Replies are listed 'Best First'.
Re^2: Issue with regex matching
by AnomalousMonk (Archbishop) on Aug 29, 2013 at 00:46 UTC
    while ($_ =~ m/\Q$constant\E/gc) { $count++; }

    Minor point: in  m/\Q$constant\E/gc the  /c modifier is unnecessary, although it does no harm in this instance...

    ... although wanna_code_perl's
        $count += () = /\Q$constant\E/g;
    in a readline loop or
        $count = () = /\Q$constant\E/g;
    on a slurped file is, IMHO, better.

Re^2: Issue with regex matching
by sowais (Sexton) on Aug 29, 2013 at 21:40 UTC
    Thanks for your input rminner. Adding the \Q and \E as you stated worked. I will try with quotemeta as well. I admit I could have done a better job in explaining the objective, I need to count the number of lines where the $constant variable appears in the file. For the points you mentioned under 'Other thoughts':
    - I do use something like $FILE in the code but to make the code more easy on the eyes I simply inserted the code where the problem was occurring.
    - Its a line counter script that I am converting from batch, and to keep things consistent for the users, I kept the same terminology. I do agree term could use a change.
    - I am doing something else if $constant is empty/undef, that part is working properly so thought there is no sense in copying all of that.
    Thanks Again!