Re: Benefits of declaring literals in D-specs

One of the reasons I remember being told to do this is what if they change
account types. The business expanded or some VP has a "wonderful new
system". Account 32 is now Account 132. In each program you only need to
update one place and recompile.





Alan Shore <AlanShore@xxxxxxxx>
Sent by: rpg400-l-bounces@xxxxxxxxxxxx
05/03/2007 09:58 AM
Please respond to
RPG programming on the AS400 / iSeries <rpg400-l@xxxxxxxxxxxx>


To
RPG programming on the AS400 / iSeries <rpg400-l@xxxxxxxxxxxx>
cc

Subject
Re: Benefits of declaring literals in D-specs






Brian - I agree with your thought process, but over time, I have found
there is one problem with this idea.
To you, Acct_type 32 is defined as DEAD-BEAT, but to someone else
Acct_type
32 is defined as DEFUNCT_TYPE, therefore a search on DEFUNCT will result
in
nothing found, and therefore a search of 32 is then required.
By itself - no great shakes, but somewhat irksome.



Alan Shore

NBTY, Inc
(631) 244-2000 ext. 5019
AShore@xxxxxxxx
"If you're going through Hell, keep going" - Winston Churchill

rpg400-l-bounces@xxxxxxxxxxxx wrote on 05/03/2007 10:30:28 AM:

Named literals avoid the magic number anti-pattern. What does these
statements mean:

if status_code = 32
return
endif

if acct_type = 32
...
endif

What the heck is 32? Are both 32's the same thing?

Compare to this

d EXIT_KEY_PRESSED C 32
d DEAD_BEAT C 32
...

if status_code = EXIT_KEY_PRESSED
return
endif


if acct_type = DEAD_BEAT
...
endif

Now the code is much clearer. And, if you want to locate code that
processes dead-beat accounts, it is ead to find. If you search for 32,

you

have to examine each reference and decide if it is account type or some
function key thing.