BBCode Parser

[Pikachu]

I've decide to code my own BBCode parser in Perl:

#!/usr/bin/perl
use strict;
use CGI qw(:standard);

our $query = new CGI;
our $_ = $query->param('text');

print "Content-Type: text/html\n\n";

s@\[b\](.*?)\[/b\]@<b>$1</b>@ig;
s@\[i\](.*?)\[/i\]@<i>$1</i>@ig;
s@\[u\](.*?)\[/u\]@<u>$1</u>@ig;
s@\[big\](.*?)\[/img\]@<span style='font-size: large'>$1</span>@ig;
s@\[url=((?![^:]+script).*?)\](.*?)\[/url\]@<a href="$1">$2</a>@ig;
s@\[url\]((?![^:]+script).*?)\[/url\]@<a href="$1">$1</a>@ig;
s@\[img\]((?![^:]+script).*?)\[/img\]@<img src="$1" alt="[IMAGE]" />@ig;
s@\[center\](.*?)\[/center\]@<p style="text-align:center">$1</p>@ig;
s@\[left\](.*?)\[/left\]@<p style="text-align:left">$1</p>@ig;

print "<form action='' method='post'><textarea name='text'></textarea><input type='submit' value='submit' /></form>";
print $_;

It works fine but when translated to HTML, it isn't parsed correctly. For example, [b][i]bold[/b][/i] becomes <b><i>bold</b></i>. How would I write the script so that it parses correctly in HTML?

 

[IIMarckus]

Why not do straight HTML parsing with a DOM? Regular expressions aren’t really suited for this sort of thing—formatting tags would probably be better interpreted with a stack.

 

[Abwayax]

It works fine but when translated to HTML, it isn't parsed correctly. For example, [b][i]bold[/b][/i] becomes <b><i>bold</b></i>. How would I write the script so that it parses correctly in HTML?

That's because you nested the tags wrong in the first place. If you open "b" before "i", you close out "i" before closing "b."

[b][i]foo[/i][/b] becomes <b><i>foo</i></b>, which is valid HTML.

 

[Pikachu]

Yeah, I know that. I want to know how to fix that using the script. It's for a guestbook.

 

[Abwayax]

Well, the way MyBB does that is that, instead of using "b" and "i" tags, it uses span tags like so

<span style="font-weight: bold;"><span style="font-style: italic;">foo</span></span>

That way, if the tags are improperly nested, it doesn't matter because they're both the same type of HTML tag. vBulletin actually rearranges the tags. I'm not sure how they do that because I don't have a vBulletin license.

 

[Pikachu]

Hmm, that work to an extent. What if they wanted a URL italicizes: [url=#][b]hi[/url][/b]

Thanks for your help, though! That made it a bit easier.