[dasdingansich]

Originally Posted by Digital Jedi

It's tricky when you run into an area you need grouped that can easily match the rest of the string. One thing you can do is check to see if the underscore in your grouping is absolutely necessary. If not, then you can easily put an optional word character match after your grouping: [\w_]*

If it is necessary, then try looking for the user name someplace else besides the URL string. Likely, it will be somewhere in the HTML of the page and you can make a ReGex match to that. In this case your grouping, the $p1 reference, will be in the Extraction regex instead.

good idea, thanks!
the username occurrs once between h1-tags on the target page, so this should do the trick:
<h1>([\.\w_-]+)</h1>

only Problem is, AME stops fetching the page content at the first occurrence of german vowel "?", which unfortunately appears right in the title.
please compare
http://shop.ebay.com/merchant/25cent-giraffes
http://shop.ebay.de/merchant/25cent-giraffes
the word "seller" in the page title becomes "Verk?ufer", and AME stops at "Verk", which in vbulletin results in this:

[ame="http://shop.ebay.de/merchant/25cent-giraffes"]eBay Verk

it?s definately the vowel because it works perfectly with ebay.com, but that?s no good to me because they?re running an entirely different affiliate program.