Extract last tagName of various Selectors

Question

Basically I attempting to extract the last tag name of a handful of different css selectors.

I have successfully implemented what I am talking about in javascript, I'm looking for a more compact way using only 1 regex expression preferably.

Here is what I successfully have working.

//Trim selector, remove [attr] selectors due to conflicts, and standardize >,~,
selector=selector.replace(/^(\s|\u00A0)+|(\s|\u00A0)+$/g,'')
selector=selector.replace(/\[[^\]]*\]/g,'[]').replace(/\s*([>~\s])\s*/g,'$1');
var theSplit = selector.split(/[>~\s]/);
selector  = /^[^.\[:]*/.exec(theSplit[theSplit.length-1]) || "*";

I am only looking to support css 2.0 selectors that are 100% supported by internet explorer 7.

For example + selector and :first-child selectors are static in ie7 and therefore I have no need to support them. Here is a list of css selectors that must work.

#test span ul a
#test >span[style="background:green"]
#id + span ~ article.class
section header.class
body
div
body div
div p
div > p
div ~ p
div[class^=exa][class$=mple]
div p a
.note
div.example
ul .tocline2
#title                                  
h1#title
div #title                                      
ul.toc li.tocline2
ul.toc > li.tocline2
a[href][lang][class]
div[class]
div[class=example]
div[class^=exa]
div[class$=mple]
div[class*=e]
div[class|=dialog]
div[class!=made_up]
div[class~=example]

Edit : I ended up using this script. It even takes into consideration the universal selector

var lastTagName = selector.replace(/\[[^\]]+\]|[\.#][\w-]+|\s*$/g,'').replace(/^.*[^\w]/, '')||'*'

Answer 1

The stuff at the beginning of the selector should be easy enough to remove as a last step, so let's focus on removing the two types of junk from the end first. That's stuff in square brackets and stuff following a . or #.

str = str.replace(/\[[^\]]+\]/g, '').replace(/[\.#][a-z0-9]+/gi, '');

That is, remove everything between square brackets (in the entire expression; it's okay if it catches stuff earlier), and then likewise for . and # selectors.

Then, finally, take out everything leading up to a non-alphanumeric character, which covers spaces, >, et cetera. This will leave only the last "word" (in this case a tag name) in the string:

str = str.replace(/^.*[^a-z0-9]/i, '');

A jsFiddle to try it out: http://jsfiddle.net/JAA98/