[comp.databases] End-of-file characters in .DBF files

jdouglas@wheaton.UUCP (John Douglass) (07/26/89)

I am writing database applications with Clipper on a Novell Netware 286
network.  I have had trouble with end-of-file characters (^Z) appearing in
the middle of a .DBF file, so that a SEEK for a record after the ^Z will fail.
The ^Z's appear in the byte preceding a record (which usually contains a
space, or '*' for deleted records).  Also, the three records before ^Z are
duplicated _after_ the ^Z like this:

		    .
		    .
	    <space>Record 323 info
	    <space>Record 324 info
	    <space>Record 325 info
	    ^ZRecord 323 info
	    <space>Record 324 info
	    <space>Record 325 info
		    .
		    .

This can be fixed using DEBUG by replacing the offending ^Z with a space,
but the 3 duplicated records must be deleted.

This problem has occurred in several different files, but is sporadic 
enough that it's hard to find the exact cause.  Any help in finding 
a solution would be appreciated.