Re: Text processing on source members

[CRPence <crp@xxxxxxxxxxxxxxxxxxxx>]

  Either of the following two [effectively identical] SQL statements 
requested, after issuing OVRDBF SS TOFILE(TheLib/TheFile) MBR(TheMbr) 
OVRSCOPE(as_required), will perform a request to "find a line containing 
X which is not followed by a line containing Y or Z."

  The RRN() is used because the increment for SRCSEQ is unknown, as is 
the uniqueness of SRCSEQ values.  If the member were known to have been 
reorganized with SRCOPT(*SEQNBR) SRCSEQ(1.00 1.00) and not since 
changed, the RRN() could be replaced with SRCSEQ.

     select rrn(r1),r1.*,rrn(r2),r2.*
     from          ss r1
      inner join   ss r2
      on  rrn(r1) = rrn(r2) - 1
      and r1.srcdta like '%X%'
     where r2.srcdta not like '%Y%'
       and r2.srcdta not like '%Z%'

     select rrn(r1),r1.*,rrn(r2),r2.*
     from          ss r1
      inner join   ss r2
      on  rrn(r1) = rrn(r2) - 1
      and r1.srcdta like '%X%'
     where locate('Y', r2.srcdta)=0
       and locate('Z', r2.srcdta)=0

  Invoked as a [an effective] function against each desired 
file.member, the select could be adjusted to return the first RRN(R1).

Regards, Chuck