One small point: if you are doing a lot of different substitutions, you may feel tempted to do something like

$text =~ s/(\W+)/$1->()/eg; # replace any nonwhitespace characters wit
+h
the result of a function call
[download]

The danger is that unless you trust the file being fed into the program, someone might feed you a word that called some bad subroutine in your program. Or it might just happen accidentally!

To make sure this doesn't happen, set up a hash of subroutine references(I think this could be called a "dispatch table"):

%subs = (
    bar => \&bar
    foo => \&foo
    do_something => \&do_something
);

$text =~ s/(\W+)/$subs{$1}->() if defined $subs{$1}/eg; # only call fr
+om a list of specific subroutines
[download]

The $subs{$1}->() is calling the subroutine reference (coderef) defined in the hash, so for example "bar" would be replaced by the result of the "bar" subroutine.

dave hj~