[translations] new web friendly extractor

[Gabor Hojtsy]

Hi,

I said in february that we got a new friendly extractor, which 
autodiscovers the Drupal files, so we don't need any command line magic. 
Today's update means that we don't need any command line at all. Just drop 
the extractor in any folder under Drupal in your webroot, access the file 
via the webserver in a browser, and see the pot files generating. The 
script still autodiscovers all Drupal related files in all subfolders 
recursively.

Grab the new file from CVS or: http://href.hu/x/1i74 (a shortened CVS 
checkout address).

Ps. actually the extractor worked over the web fine already, it just 
printed an ugly line and then nothing. This was changed today. It still 
prints all the t() errors found, so those can be fixed.

Gabor