>>>I'm using a query which uses the set filter to command. Good so far. But when the result of the set filter is no records, I need to be able to test for that, so I can take some action (message or otherwise). I see that RECC() doesn't respect the Set Filter to command. Can anyone tell me what does?
>>>
>>>Thanks!
>>>
>>>Sylvia
>>

function filterhasrecords
>>locate
>>return !eof()

Cetin
>
>Great thanks! By the way, I also tried a simple count to l_nreccount which returned a 0, and then I tested for this value. Is this method going to be slower than locate?
>
>Sylvia
Sylvia,
When expression would lead to 0 recs, I think count, locate, go top etc would nearly be equal. But I would still suggest locate if you use set filter for one reason. Supposing your expression would waste some time to filter, if result is not 0 records then you would set filter after count which would double the time (in fact less than double but still more than one pass). On 0 records situation clearing a filter is fast. Another approach on large table (and buffering) is to do an SQL to a true cursor, do edits in that cursor, and write changes to table (faster on large tables - drammatically where filter not optimizable).
Cetin