Basically I want to remove any underscores that have been coded into CSS classes, so :

.test_left

Would become

.testleft

But would leave other underscore untouched, for instance href links with _blank targets and the like, so as long as the underscore is preceeded by a word and a period

find a period (followed by zero or more word characters) followed by an underscore (followed by 0 or more of any character matched pessimistically) followed by a word boundary

replace with a period $1 $2



you might have to repeatedly run this replacement as it only removes the first underscore in any identifier

\.(\w*)_(.*?)\b

.$1$2



try that...

Hey there, thanks for the tip, unfortunatley this doesn't even find a string. I tried with UNIX both on and off just to ensure that wasn't affecting anythihg, but didn't seem to work. If you have a solution using UltraEdit RegEx that would be superb!

Thanks for the aisstance anyways!

Hi

Using the Perl regular expression engine you would do the following:

Find what: .(\w+)_(\w+)

Replace with

.$1$2

cheers

Hey there,

I must be doing something wrong because none of those solutions work unfortunately

Thanks all for the help tho!

Cheers

You're probably not doing anything wrong. I think the reason those suggested solutions don't work is because "\w" matches underscores (it's equivalent to "[a-zA-Z0-9_]"). Also the period needs to be escaped because it normally means "any single character".

Find:
(\.[a-zA-Z][-a-zA-Z0-9]*)_([-a-zA-Z0-9]+)(\b|_)

Replace:
$1$2

That should remove up to two underscores from class names. For classes with more than two underscores you'll have to run the search/replace again.