[daip] SNSMO bug

[Eric Greisen]

Michael Bietenholz writes:
 > SNSMO dies for no apparent reason when the last record is flagged.
 > Problem is in RFFXFG - it's because the IRET=-1 from TABIO gets kept
 > and returned as the IRET from RFFXFG, even though RFFXFG ended fine.
 > A simple fix below (this is DEC02 version; but seems like DEC03 version
 > has same problem but the LOOPR label is 20 not 100)

I have put the suggested fix many places in SNSMO.  More disturbing was
the fact that many DO 200 loops did not test for flagged rows and
rewrote them as unflagged rows.  I fixed that too.  Tomorrow's MNJ
should have the fix.

Thanks,

Eric Greisen