Loren:
>From you description it appears that the contents of each record is uniform.
If that is the case, it should be straight forward, if tedious, to implement
an algorithm to estimate the number of records from the file size, and the
compiler manual's description of its I/O implementation, e.g., if a typical
implementation the size of integer used to define the number of words/bytes in
each record. The algorithm would be system dependent but easily isolated.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|