Alternative to FOUND_ROWS()

[jun kazawa]

FOUND_ROWS is convenient when you have to get the count of all rows from a SELECT SQL_CALC_FOUND_ROWS that contains a LIMIT clause.

However we need the count(*) of a query from a huge table where we want the count to be limited to a given number (e.g. 1000). It is almost the opposite of FOUND_ROWS().

We want actually to take advantage of the indexes and the LIMIT algorithm that reduce the number of rows examined, without returning the actual rows - just the count. The count would be actually and virtually MIN(Limit_given, count(*))

Is it possible to get such a limited count(*)?