Date:         Fri, 24 Oct 1997 01:14:34 GMT
Reply-To:     Jan Padilla <jpadilla@HOOKED.NET>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Jan Padilla <jpadilla@HOOKED.NET>
Organization: Netcom
Subject:      Re: Help: Flatfile update, without reading the data into SAS

On Thu, 23 Oct 1997 17:40:34 -0700, Scott Carl <SCARL@THECREEK.COM>
wrote:
 
>
>I realize this is a simple join using either SQL or datastep processing,
>but since the objective is to replace the flatfile with an updated
>flatfile, I don't think reading the data into SAS datasets, joining the
>two datasets and writing the data back to ASCII will be efficient. Any
>thoughts will be appreciated.
 
Hi, Scott,
 
I you know any other programming languanges, this would probably be a
good time to use them...COBOL or C, for example, can handle a table
lookup much more efficiently.
 
If that's not an option, you could set up the small dataset as a VSAM
file if you're on a mainframe, or I've been known to use a MACRO to
build one big if statement, like:
;
DATA _NULL_;
  INFILE INFILE:
  INPUT VAR1;
  IF VAR1 = OldVal1
    THEN PUT OldVal1;
  ELSE IF
 VAR1 = OldVal2
    THEN PUT OldVal2;
   ...
 ELSE IF Var1 = OldValN
   THEN PUT OldValN;
 ELSE PUT _INFILE_;