Regular expression help: Taking HTML-like attributes with optional quotes

Cap'n Steve has asked for the wisdom of the Perl Monks concerning the following question:

I'm currently writing something that uses a variant of bbcode to allow the user to create an HTML form. It works so far, but in an effort to idiot-proof it, I've decided to make putting quotes around the attributes optional. This is the regexp that does that:

/name=(")? # "name" attribute followed by optional quote

([^(?(1)"|\s\])]+) # if the quotation mark was present, match until an
+other one, otherwise stop at whitespace or a bracket

(?(1)") # match end quote, if applicable
/xi
[download]

What I'm wondering now is if I should allow users to escape quotes in the attribute. I know this is more trouble than it's worth, but I'm curious. How would I go about ignoring the quotation marks preceded by a backslash? I've tried a negative lookbehind within the conditional, but it didn't work.

Comment on Regular expression help: Taking HTML-like attributes with optional quotes Download Code

Replies are listed 'Best First'.
Re: Regular expression help: Taking HTML-like attributes with optional quotes by tlm (Prior) on Jun 28, 2005 at 10:24 UTC
Take a look at Text::Balanced. the lowliest monk	[reply]
Re: Regular expression help: Taking HTML-like attributes with optional quotes by rlucas (Scribe) on Jun 28, 2005 at 07:08 UTC
... might be one for Parse::RecDescent. If you need something performant, P::RD may not be the way to go. But if you've never programmed a parser before, using P::RD will make you a better programmer for the experience. Also, Parse::RecDescent is very well documented. A final bit of info to consider is that there is a set of regexes that Perl will not play nice with (due to how the regex engine works, being nondeterministic, IIRC). By "not play nice," I mean take billions of CPU cycles for ostensibly simple decisions. I think that dealing with optional escaped quoting with regexes might be putting you into that territory.	[reply]