[Resolved] "BBCode"-style formatting with regexps.

[gazchap]

I'm creating a little CMS for work, and am allowing users to use "BBCode" style tags to format their text.

I am allowing them to insert (very) basic tables too, using a set of [ TABLE ], [ ROW ], and [ COL ] tags.

However, I'm also converting line breaks in the text to <br />, so if someone types in this:

[ TABLE ]
[ ROW ]
[ COL ]Hello[ /COL ]
[ /ROW ]
[ /TABLE ]

Then the table ends up halfway down the page because of the <br /> that gets put at the end of each line.

How can I trap these in the regular expression? The regular expression I'm using so far is:

	"\\[TABLE\\](.*?)\\[\/TABLE\\]" => "<table border=\"0\" cellspacing=\\"4\\" cellpadding=\\"4\\">\\\\1</table>",
	"\\[ROW\\](.*?)\[\\/ROW\\]" => "<tr>\\\\1</tr>",
	"\\[COL\\](.*?)\[\\/COL\\]" => "<td>\\\\1</td>",

(using arrays to do a preg_replace, so the search pattern is in the array key, and the replace pattern in the keys value)

I've tried using \n* but it doesn't seem to match.

Can anyone help?

[webdes03]

Interesting delima... what if you replace the "[table][row][col]" string before you process the line breaks. Run the conversion twice, doing one part each time... I would think if you formatted the tables first, then it wouldn't pick up the extra line breaks... don't know if it'll work though.

[gazchap]

Cheers, don't know why I didn't think of that before.

Of course doing a regexp to match a \n isn't going to work after I've just converted it to a <br /> 😃

God, I feel stupid now 😉

One thing to note if any other people come across this problem and find this thread - if you're developing on a Windows system, make sure you either strip out the carriage returns (\r) or catch them in your regular expression as well, otherwise it won't work.

My regular expression for matching the table code looks like this:

\[TABLE\]\r*\n*(.*?)\[\/TABLE\]\r*\n*