RE: CHAIN to a DS

[Marco Facchinetti <facchinetti@xxxxxxxxx>]

About point 4: our standard approach to avoid moving data
from EQUAL files is this:

ffilea     if   e           k disk    prefix(A_)           
   
ffileb     uf a e           k disk    prefix(B_)           
   
ddsfilea        e ds                  extname(filea)
prefix(A_)
ddsfileb        e ds                  extname(filea)
prefix(B_)
 /Free              
  chain (key1:key2) filea;   / Now dsfilea is filled from
filea
  chain (key1:key2) fileb;   / Now dsfileb is filled from
fileb
  dsfileb=dsfilea; /Now dsfileb is equal to dsfilea...     
   
 /End-free          

If the library name is at the top or bottom of the record
using evalr ans %subst should be enough. Mind %NULL
values...

HTH
Marco     
--- CarollaT@xxxxxxxxxxxxxx wrote:

> Since writing this, the strategy has changed, so I am
> doing the following:
>  
> 1.        Read journal entries from several TRANS files
> (all exact same
> format, different libraries) into an output file (using
> DSPJRN - I know,
> Scott -- I should be using an API ;)
> 2.           Read through each journal entry, and place
> the record image
> field (JOESD) into an externally-described DS, based on
> the TRANS file
> (minus one field)
> 3.           Attempt a chain to the Master TRANS file
> (fields prefixed with
> a 'D')
> 4.           Place all fields of data from the DS into
> the Master TRANS file
> 5.           Update or Write the Master TRANS file, based
> on the results of
> the CHAIN op
>  
> I just wanted to know if there was a way to avoid step
> four - because there
> are a lot of fields, and I want this to run fast Fast
> FAST.  If I avoid
> prefixing the Master TRANS file fields, then the fields
> in the external DS
> would be the same fields, right?  So after I populate the
> DS with the
> journal image, I want to position myself to do a write or
> update to the
> Master TRANS file, without replacing the contents of
> these fields.  I am
> pretty sure you can't do an UPDATE after doing a SETLL,
> otherwise, I could
> do a SETLL, then check the %EQUAL bif, and optionally
> update or write.  
>  
> SQL is not _really_ an option, because we don't have the
> development LP.  I
> know there is still a way to interact using SQL, but I
> haven't really 'gone
> there' yet.  
> 
>  
> 
> Tony Carolla
> 
> MedAmerica Billing Services, Inc.



                
_______________________________
Do you Yahoo!?
Win 1 of 4,000 free domain names from Yahoo! Enter now.
http://promotions.yahoo.com/goldrush