http://www.perlmonks.org?node_id=211046

One way to change URIs in Text to HTML-Links

Use the module URI::Find or URI::Find::Schemeless, e.g

<Update>
Added encode_entities in the following code because of merlyn's answer (Thank you very much!)
</Update>

#! /usr/bin/perl use strict; use warnings; use URI::Find::Schemeless; use HTML::Entities qw(encode_entities); # changed my $text = q~ hello this is no.url this is an url: www.fabiani.net ftp.anything.de/test/thisfile mailto:martin@fabiani.net or the like yeah martin@fabiani.net http://www.fabiani.net/ ~; # create a new URI::Find::Schemeless objekt and add as callback # the function what shell be done with each found URI my $finder = URI::Find::Schemeless->new ( sub { my ($uri, $originalUri) = @_; # error: encode_entities is missing # return qq~<a href="$uri" target="_newpage">$originalUri</a>~; return q/<a href="/ . encode_entities("$uri") . q/">/ . encode_entities($originalUri) . q/>/; } ); # here starts the search (and in our case the replacement): my $howManyFound = $finder->find(\$text); # lets have a look at the result print "$howManyFound URIs found\n"; print "$text\n";
This will replace the following URIs:
  • www.fabiani.net
  • ftp.anything.de/test/thisfile
  • mailto:martin@fabiani.net
  • http://www.fabiani.net/
If you just want to replace the following URIs, use URI::Find, which is more strict:
  • mailto:martin@fabiani.net
  • http://www.fabiani.net/
You can do this by killing ::Schemeless:
use URI::Find; # instead of URI::Find::Schemeless ... my $finder = URI::Find->new # instead of URI::Find::Schemeless ( sub { my ($uri, $originalUri) = @_; # error: encode_entities is missing # return qq~<a href="$uri" target="_newpage">$originalUri</a>~; return q/<a href="/ . encode_entities("$uri") . q/">/ . encode_entities($originalUri) . q/>/; } ); ...
If you dont want the Links to open a browser in a new window, just kill target="_newpage"

It is just a shame that these modules are not standard modules of perl, but I hope that they soon will become.

If your provider hasn't installed them and doesn't want to, just copy the directorries of URI to your webpath, e.g. to cgi-bin/lib and load them perhaps from your cgi-scripts which are located in cgi-bin with the modules FindBin and lib:

BEGIN { use FindBin qw($Bin); use lib "$Bin/lib"; } use URI::Find::Schemeless;

Big thanks to mdupont for pointing me to the new interface of URI::Find (was working with find_uris for a long time, and with a piece of code much too complicated)

Best regards,

strat

Replies are listed 'Best First'.
•Re: Change URIs in Text to HTML-Links
by merlyn (Sage) on Nov 07, 2002 at 16:37 UTC
    You left out the mandatory "escape HTML entity" calls. Shame on you.
    use HTML::Entities qw(encode_entities); use URI::Find; ... my $finder = URI::Find->new # instead of URI::Find::Schemeless ( sub { my ($uri, $originalUri) = @_; return join "", q/<a href="/, encode_entities("$uri"), q/">/, encode_entities($originalUri), q/>/; } ); ...
    Yes, I've already complained to the author. He "fixed" the manpage incorrectly, by escaping the entire source text first. That breaks in-page URLs that look like http://example.com/foo/bar?a=b&c=d. {sigh} I was too worn out explaining to him why that was wrong to submit another fix.

    -- Randal L. Schwartz, Perl hacker
    Be sure to read my standard disclaimer if this is a reply.

Re: Change URIs in Text to HTML-Links
by Aristotle (Chancellor) on Nov 08, 2002 at 01:30 UTC
Re: Change URIs in Text to HTML-Links
by Anonymous Monk on Jan 01, 2004 at 14:44 UTC
    The problem with URI::Find is, that it locates URI's inside href tags. URI::Find is alright if your not allowing HTML to be parsed.