Looking for regexp - [list] items

[eluser]überfuzz[/eluser]
I'm looking for regular expressions to use in a bbcode parser. My current parser is working ok, but I'm not good enough with regexp to muster good expressions to handle

• ; ordered, unordered, decimal, etc. It seems that I always miss out on something...
  
  I would be very thankful if anyone here could point me in the right direction, where as to look for them. Or if anyone has a regexp that can handle nestled
  • with or without different list-style.
    
    Thanks in advance.

[eluser]WanWizard[/eluser]
In our old forum code we use

Code:

// old style lists
$text = preg_replace('#\[li\](.*?)\[/li\]#si', '<li style=\'margin-left:35px;\'>\1</li>', $text);
$text = preg_replace('#\[ul\](.*?)\[/ul\]#si', '<ul style=\'margin-left:-15px;\'>\1</ul>', $text);

// new style lists
$text = preg_replace('#\[list=1\](.*?)\[/list\]#si', '<ol>\1</ol>', $text);
$text = preg_replace('#\[list\](.*?)\[/list\]#si', '<ul>\1</ul>', $text);
$text = preg_replace('#\r\n\[\*\]#si', '<li>', $text);

[eluser]überfuzz[/eluser]
Thats about what I can handle myself. I'm running into trouble if I'm trying to extend your regexp:es. First of all I can't extend them to handle upper-roman nestled with lower roman. Furthermore, I can't get the HTML output to validate. Shouldn't the nestled <ol>list items</ol> sit inside the <li>here</li> tags?

Example of how I would like to extend it, not working but still...

Code:

$string = preg_replace('#\[list=a\](.*?)\[/list\]#si', '<ol style="list-style-type:lower-alpha;">\1</ol>', $string);
    $string = preg_replace('#\[list=A\](.*?)\[/list\]#si', '<ol style="list-style-type:upper-alpha;">\1</ol>', $string);
    $string = preg_replace('#\[list=I\](.*?)\[/list\]#si', '<ol style="list-style-type:upper-roman;">\1</ol>', $string);
    $string = preg_replace('#\[list=i\](.*?)\[/list\]#si', '<ol style="list-style-type:lower-roman;">\1</ol>', $string);

[eluser]WanWizard[/eluser]
Issue with the upper/lower is probably that you use a case-insensitive regex, so list=I equals list=i, which makes them all upper.

[eluser]überfuzz[/eluser]
[quote author="WanWizard" date="1273578687"]Issue with the upper/lower is probably that you use a case-insensitive regex, so list=I equals list=i, which makes them all upper.[/quote]

Wops, I see that I missed out on posting an example of the output. The upper/lower roman chars were mixed, seemingly in a random matter, if I nestle the lists.