Re: DISPLAY_JOURNAL and ENTRY_DATA parsing using pure SQL

[Sam_L <lennon_s_j@xxxxxxxxxxx>]

The EXPJRNE command is worth looking at. You specify a file and it gives 
you most (probably all) selections options available on the DSPJRN 
command. And the output shows the columns in the file being journaled,
It is freeware. See: 
http://www.tools400.de/English/Freeware/Utilities/utilities.html
However, it does create an output file, which is really easy to query 
with SQL.

But...
The author (Thomas Raddatz) is also the author of the iSphere plugin for 
RDi.  It has a Journal Explorer tab, but it interprets a file created by 
a DSPJRN execution. So you still have the issue of IO.
http://isphere.sourceforge.net/help/html/journalexplorer/journalexplorer.html

I don't know if Thomas would consider a request to create a UDF. He is 
normally very open to requests, but he also has a real job.

FWIW: I don't know much you expect to use the journal entries, but 
unless you have massive amounts of data to process, I'm not sure it will 
make much of an impact.

But if it is, then a UDTF may be the answer, and Scott Klement's article 
will be useful.

Sam

On 5/1/2019 2:07 PM, Darren Strong wrote:
> I'm not returning a single row, and, the bigger issue is that the columns vary for every journal file you parse. It looks to me like you have to define the result set up front for a UDTF, and the whole point of this issue, and why I suspect IBM has avoided this thing, is that its so dynamic.