RE: Mass conversion of MOVE to EVAL.

["Tyler, Matt" <mattt@xxxxxxxxxxxxxx>]

Carel,
        I do not think that is a good solution since MOVE "only" moves the
source characters to the target overlaying only the length of the source
data.  EVALR is a clear and then right adjust.  The solution that Linoma's
convert RPG source tool gives is much better.  It %SUBST() the target with a
start and length appropriate of the source field.  

        Now replacing MOVE with EVALR is okay if you again %SUBST() so that
you are only replacing those specific positions.

Thank you,
Matt Tyler
WinCo Foods, LLC
mattt@xxxxxxxxxxxxxx


-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx [mailto:rpg400-l-bounces@xxxxxxxxxxxx]
On Behalf Of Carel Teijgeler
Sent: Wednesday, August 10, 2005 12:49 PM
To: rpg400-l@xxxxxxxxxxxx
Subject: Re: Mass conversion of MOVE to EVAL.

Rob,

I think MOVE should be replaced with EVALR, as EVAL is similar to MOVEL.

Just a little reminder before you make erronuous conversions.

Regards,
Carel Teijgeler

*********** REPLY SEPARATOR  ***********

On 9-8-05 at 15:55 rob@xxxxxxxxx wrote:

>If you suspect that the only use of MOVE was the fact the gent started
coding before EVAL existed,
>and, there were no 'tricks' done with MOVE,
>what do you think about doing a mass search&replace on MOVE with EVAL?  If 
>I immediately recompile and it goes successfully what problems would a 
>hidden trick play that the compiler wouldn't catch?  For example, moving 
>alpha to numeric, or vice versa, would blow the compile, right?  Then I 
>would know that I need to handle that on a individual basis.
>
>MOVE CHAR15  CHAR4 sure behaves differently than EVAL CHAR4=CHAR15
>Would this be flagged on a compile listing as a low level warning?  And
vice versa flagged also?
>(Just tried it.) Answer is no.
>
>Any compiler option that will have it generate a message?
>
>Goal:  Convert all these silly MOVEs to EVALs as a first step to converting
the source to free format.