Date:         Fri, 9 Nov 2007 22:50:52 -0500
Reply-To:     Steve First <sfirst@SYS-SEMINAR.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Steve First <sfirst@SYS-SEMINAR.COM>
Subject:      file compare utility

Hi,
I am looking for a more robust file compare routine than proc compare.  What
I would like is a sas technique that works like the unix DIFF command.
I have two SAS ds, where each row may match or it may have extra rows on
either ds.  There is no common key or sort order.

Example:

  ds1      ds2

   x        x

   1        1
   2        3
   3        4
   5        5
            5
            4

What i would like would be something like.

   x     x_1

   1     1
   2     .
   3     3
   .     4
   5     5
   .     5
   .     4




Proc compare doesnt seem to handle extra rows in either of the two files
unless there is a key (id).  There is some written about the diff algorithm,
but they are mostly written in c or such, and Im a SAS guy.
Does anyone have an algorithm that can do this and be adapted to SAS?

Thanks,

Steve