On Sun, Oct 05, 2008 at 05:22:37PM +0700, Henrik Sarvell wrote: > (class +Awords +Entity) > (rel words (+Key +List)) > (rel counts (+List)) > ... > [idxtest.l:2] !? (extra Var (cdr Lst)) > T -- Bad extra
Oops, sorry, my fault! It should have been (rel words (+List +Key +String)) (rel counts (+List +Number)) There were only prefix classes, but no main relation class, so it can indeed not find any 'extra' ;-) BTW, with (+List +Key +String) it will create an index entry for each word in the list, while with (+Key +List +String) the index will be the whole list. Both cases don't make much sense, IMHO. Do you really want to index all those words? If yes, perhaps better to use a non-unique index with (+List +Ref +String)? Cheers, - Alex -- UNSUBSCRIBE: mailto:[EMAIL PROTECTED]
