[tpm] Search/Replace multiple patterns in a single line, in a single pass?

[J. Bobby Lopez]

I omitted the /g on the regex in the sample code, but the working code has
it.

The working code would deal with more complex patterns using a combination
of letters, numbers and special characters, for example MAC/IP addresses,
hostnames, etc.  It could be any pattern of characters really, but they
would have to be replaced in the same way that Mike suggested, where the
surrounding data is not affected.



On Wed, Dec 22, 2010 at 12:54 PM, J Z Tam <jztam at yahoo.com> wrote:

> I'm with Richard:
> Why not just tr the line?
>   DB<1>
> main::(bobbyRegex1.pl:10):      my $line = "*!*!* a *!*!* b *!*!* c *!*!*";
>
>   DB<1>
> main::(bobbyRegex1.pl:14):      my $pattern = '(a)+\s.*(b)+\s.*(c)+\s'; #
> kinda/
>
> sorta, replaces stuff between patterns also
>
>   DB<1> $expectedMatches = $line =~ tr/abc/def/;
>
>   DB<3> x $expectedMatches
> 0  3
>
>   DB<4> x $line
> 0  '*!*!* d *!*!* e *!*!* f *!*!*'
>
>   DB<5>
>
> --- On *Wed, 12/22/10, Richard Dice <richard.dice at gmail.com>* wrote:
>
>
> From: Richard Dice <richard.dice at gmail.com>
> Subject: Re: [tpm] Search/Replace multiple patterns in a single line, in a
> single pass?
> To: "J. Bobby Lopez" <jbl at jbldata.com>
> Cc: "Toronto Perl Mongers" <tpm at to.pm.org>
> Received: Wednesday, December 22, 2010, 12:20 PM
>
>
> Two thoughts.
>
> Why aren't you using the 'global' switch on the substitution, like s///g ?
>
> Second, if the substitutions in your real code are as simple as in your
> example code, you should consider the transliterate operator rather than the
> substitution operator, per $foo =~ tr///;
>
> Sent from my iPhone
>
> On 2010-12-22, at 10:44 AM, "J. Bobby Lopez" <jbl at jbldata.com<http://mc/compose?to=jbl@jbldata.com>>
> wrote:
>
> > Hey all,
> >
> > I'm trying to figure something out, but having some trouble.  I'm trying
> to replace multiple patterns via capture buffers, with multiple replacements
> contained in a hash.
> >
> > Looking at the code below, you can see that the patterns I'd like to
> match are contained in $regex just fine, however I'd like to have mutiple
> $replacement{$#} 's depending on the number of patterns I'm trying to
> replace.
> >
> > ==== begin-code ===
> > #!/usr/bin/env perl
> > use strict;
> > use warnings;
> >
> > my %replacement;
> > $replacement{'a'} = 'd';
> > $replacement{'b'} = 'e';
> > $replacement{'c'} = 'f';
> >
> > my $line = "*!*!* a *!*!* b *!*!* c *!*!*";
> >
> > #my $pattern = '(a)|(b)|(c)'; # Doesn't work, only matches first pattern
> > my $pattern = '(a).*(b).*(c)'; # kinda/sorta, replaces stuff between
> patterns also
> >
> > #$line =~ s/$pattern/$replacement{$1}/; # <-- here's where I'm having
> difficulty
> > $line =~ s/$pattern/$replacement{$1} $replacement{$2} $replacement{$3}/;
> # kinda/sorta
> >
> > #$line =~ m/$pattern/;
> > #print "$1, $2, $3\n";
> >
> > print $line."\n";
> > ==== end-code ===
> >
> >
> > Either I'm missing something, or this isn't actually possible, and I'll
> have to loop through the patterns and replace them individually.
> >
> > Thoughts?
> >
> > -Bobby
> > _______________________________________________
> > toronto-pm mailing list
> > toronto-pm at pm.org <http://mc/compose?to=toronto-pm@pm.org>
> > http://mail.pm.org/mailman/listinfo/toronto-pm
> _______________________________________________
> toronto-pm mailing list
> toronto-pm at pm.org <http://mc/compose?to=toronto-pm@pm.org>
> http://mail.pm.org/mailman/listinfo/toronto-pm
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.pm.org/pipermail/toronto-pm/attachments/20101222/eccc4d42/attachment.html>