RE: retrieving PF records from journal

Joel--

Here's a technique we've used here for a lo-o-ong time.

Journal entries consist of a header, which contains lib/file/mbr info, a sequence number, and other journal housekeeping stuff. The header is followed by the record image.

One way to retrieve the data in a usable form is to create a custom file for each original file you want to retrieve. File format is:


A* This is the DDS for the Special File
A* The first 19 fields make up the journal entry header.
A R RF$OH
A JOENTL 5S 0
A JOSEQN 10S 0
A JOCODE 1A
A JOENTT 2A
A JODATE 6A
A JOTIME 6S 0
A JOJOB 10A
A JOUSER 10A
A JONBR 6S 0
A JOPGM 10A
A JOOBJ 10A
A JOLIB 10A
A JOMBR 10A
A JOCTRR 10S 0
A JOFLAG 1A
A JOCCID 10S 0
A JOUSPF 10A
A JOSYNM 8A
A JORES 20A
A* The fields up to this point are the same for any file being retrieved
A*
A* The fields from here down are an exact copy of the
A* fields in the original DDS that created the file
A OHDIV R REFFLD(DIV)
A OHOTYP R REFFLD(OTYP)
A OHOVSW R REFFLD(OVSW)
A OHOVBW R REFFLD(OVBW)
A OHCRGN R REFFLD(CRGN)
A OHCNTY R REFFLD(CNTY)
A OHST R REFFLD(ST)
A OHZIP R REFFLD(ZIP)
A . . . . . . .


Then perform the following steps:

DSPJRN JRN(&JRNL/&JRN) FILE((&LIB/&FILE &FILE)) +
RCVRNG(*CURCHAIN) OUTPUT(*OUTFILE) +
OUTFILFMT(*TYPE2) OUTFILE(&OFL/&OF) +
ENTDTALEN(*CALC)

CPYF FROMFILE(&OFL/&OF) TOFILE(&OFPFL/&OFPF) +
TOMBR(&TOMBR) MBROPT(*REPLACE) FMTOPT(*NOCHK) +
CRTFILE(*YES)


&JRNL/&JRN is the journal library + journal name.

The &OFL/&OF is the library and a work file that gets created by the DSPJRN command.

The &OFPFL/&OFPF is the special file you just created to receive journal transactions for this file. It must use FMTOPT(*NOCHK) since you're taking the raw data from the journal receiver and stuffing it into the 'real' fields.

Only possible issue-- sometimes date fields don't like being stuffed. IIRC, declaring them as *TXT in the special file will let the command complete.

Paul E Musselman
PaulMmn@xxxxxxxxxxxxxxxxxxxx