Hello! I have a question that I hope the DFSORT gurus can answer. I am trying
to do a 2-file match, but there is a twist here that I am not sure I can reach
with the Sort JOINKEYS process. From what I have read, I can show matched items
or unmatched items, but not items greater than a key, while skipping those less
than or equal. I can do this programmatically, but I am already using SORT to
create the two files, so if we can stay with the theme here, that would be
great.
Maybe an example will help. My first file (F1) from a customer contains a list
of load modules with the latest maintenance APAR applied. it looks something
like this (sorry, I don't know how to use fixed-width fonts here!):
....+....10...+....20
DBCFMPR 151 LU00071
DBCL1PR 151 LU01081
DBCMPPR 151
DBCNSPR 151 LU01224
DBDMPPR 151
DBDSMPR 151 LU02807
Then, my "Master List" of APARS (F2) looks like this (there are 1 to n entries
for each module, and a lot more of these than shown)
----+----1----+----2----+----3----+---
DBCFMPR 151 RO90903 20160707 2016189
DBCFMPR 151 RO91404 20160808 2016221
DBCFMPR 151 RO94085 20170127 2017027
DBCFMPR 151 RO94783 20170329 2017088
DBCFMPR 151 RO95548 20170414 2017104
DBCFMPR 151 RO95719 20170531 2017151
DBCFMPR 151 RO99532 20171214 2017348
DBCFMPR 151 RO99997 20180308 2018067
DBCFMPR 151 SO09143 20190715 2019196
DBCFMPR 151 SO09510 20190809 2019221
DBCFMPR 151 LU00071 20210128 2021028
DBCFMPR 151 LU07865 20221115 2022319
DBCFMPR 151 LU09582 20230412 2023102
DBCGSPR 151 SO12776 20200428 2020119
DBCGSPR 151 LU00053 20210121 2021021
DBCLKPR 151 RO91115 20160726 2016208
DBCL1PR 151 SO15660 20210407 2021097
DBCL1PR 151 LU01081 20210622 2021173
DBCL1PR 151 LU04745 20220415 2022105
DBCL1PR 151 LU05861 20220727 2022208
DBCL1PR 151 LU07115 20221011 2022284
DBCNSPR 151 LU01224 20210518 2021138
DBDMPPR 151 LU07309 20221207 2022341
DBDSMPR 151 LU02338 20210913 2021256
DBDSMPR 151 LU02807 20210921 2021264
DBDSMPR 151 LU09582 20230412 2023102
Now the first list from F1 will be the driver for the process, and there are a
few scenarios:
1. F1 member/APAR is found as the latest (last, here) entry in F2;
2. F1 member/APAR is found in the list, but there are other entries later
(lower) in F2;
3. F1 member has no APAR (DBCMPPR and DBDMPPR) and there are no entries for the
member in F2;
4. F1 member has no APAR (DBCMPPR and DBDMPPR) and there are one or more
entries for the member in F2;
What I would like to see for output is:
1. Member, Version (151), APAR, "CURRENT"
2. Member, Version, APAR, "MISSING", 1st newer APAR from F2, then on
subsequent lines, each newer APAR (Member, Version, and APAR are optional on
2nd + lines)
3. Member, Version, "BASE", "CURRENT"
4. Member, Version, "BASE" , then as #2 with all missing APARs
Given F1 above, it would look like this:
----+----1----+----2----+----3----+----4
DBCFMPR 151 LU00071 MISSING LU07865
MISSING LU09582
DBCL1PR 151 LU01081 MISSING LU04745
MISSING LU05861
MISSING LU07115
DBCMPPR 151 BASE CURRENT
DBCNSPR 151 LU01224 CURRENT
DBDMPPR 151 BASE MISSING LU07309
DBDSMPR 151 LU02807 MISSING LU09582
I am not married to this output style, so if there is another way to get this
detail, we can do that. I might want to add the dates of the missing LPARs,
too, for additional clarity.
As I said, I am not sure that SORT can do this, even though it can do an awful
lot of cool things. Let me know what you think.
Thanks for your guidance and assistance!
Don Johnson
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
