Find uris which may be wrapped in enclosing delimiters.
Works like URI::Find, but is prepared for URIs in your text to be wrapped in a pair of delimiters and optionally have a title. This will be useful for processing text that already has some minimal markup in it, like bulletin board posts or wiki text.
my $finder = URI::Find::Delimited->new; my $text = "This is a [http://the.earth.li/ titled link]."; $finder->find(\$text); print $text;
my $finder = URI::Find::Delimited->new( callback => \&callback, delimiter_re => [ '\[', '\]' ], ignore_quoted => 1 # defaults to 0 ); All arguments are optional; defaults are provided (see below). Creates a new URI::Find::Delimited object. This object works similarly to a URI::Find object, but as well as just looking for URIs it is also aware of the concept of a wrapped, titled \s-1URI\s0. These look something like [http://foo.com/ the foo website] where:
The URI::Find::Delimited object will extract each of these parts separately and pass them to your callback.
\*(C`callback\*(C' is a function which is called on each \s-1URI\s0 found. It is passed five arguments: the opening delimiter (if found), the closing delimiter (if found), the \s-1URI\s0, the title (if found), and any whitespace found between the \s-1URI\s0 and title. The return value of the callback will replace the original \s-1URI\s0 in the text. If you do not supply your own callback, the object will create a default one which will put your URIs in 'a href' tags using the \s-1URI\s0 for the target and the title for the link text. If no title is provided for a \s-1URI\s0 then the \s-1URI\s0 itself will be used as the title. If the delimiters aren't balanced (eg if the opening one is present but no closing one is found) then the \s-1URI\s0 is treated as not being wrapped. Note: the default callback will not remove the delimiters from the text. It should be simple enough to write your own callback to remove them, based on the one in the source, if that's what you want. In fact there's an example in this distribution, in \*(C`t/delimited.t\*(C'.
The \*(C`delimiter_re\*(C' parameter is optional. If you do supply it then it should be a ref to an array containing two regexes. It defaults to using single square brackets as the delimiters. Don't use capturing groupings \*(C`( )\*(C' in your delimiters or things will break. Use non-capturing \*(C`(?: )\*(C' instead.
If the \*(C`ignore_quoted\*(C' parameter is supplied and set to a true value, then any URIs immediately preceded with a double-quote character will not be matched, ie your callback will not be executed for them and they'll be treated just as normal text. This is kinda lame but it's in here because I need to be able to ignore things like <img src="http://foo.com/bar.gif"> A better implementation may happen at some point.
URI::Find.
Kake Pugh ([email protected]).
Copyright (C) 2003 Kake Pugh. All Rights Reserved. This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.
Tim Bagot helped me stop faffing over the name, by pointing out that \s-1RFC\s0 2396 Appendix E uses \*(L"delimited\*(R". Dave Hinton helped me fix the regex to make it work for delimited URIs with no title. Nick Cleaton helped me make \*(C`ignore_quoted\*(C' work. Some of the code was taken from URI::Find.