Date:         Thu, 12 Feb 2004 13:05:32 -0500
Reply-To:     Shannon Fair <s.fair@CRIFGROUP.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Shannon Fair <s.fair@CRIFGROUP.COM>
Subject:      Parsing Variable Length File
Content-Type: text/plain; charset=ISO-8859-1

I am trying to read a variable length flat file with both fixed and
variable length segments.  Reading each variable length segment requires
searching for and finding each field’s tag and length in order to identify
the value in those fields.  For example, to begin reading the first
segment, I have to find a 2-byte tag (LN) of the first field (Last Name) in
that segment.  Immediately following this tag is the 2-byte length of the
Last Name field.  Since this field contains a name of eight bytes, the
record would read LN08 followed by an 8-byte character: LN08FRANKLIN.  Here
is a short example of the first part of one record:

XXXXXXXXXXXXXXXUSYYYYYYYYYY12LN04FAIRMN07SHANNON......

data test (DROP=LENGTHPN LENGTHMN);
 infile 'e:\above file.txt' missover;
 input         User_ref $15.
   Cntry_cd $2.
   Memb_cd  $10.
   Prim_cd  1.
   Sec_cd  1.
   /*Name Segment*/
   PN  $2.
   lengthPN 2. @;
   call symput ('format',lengthPN); /* I know this
doesn't work, but I would like it to*/
 input          Paternal_Name $&format..
   MN     $2.
   lengthMN  2. @;
   call symput ('format',lengthMN); /* I know this
doesn't work, but I would like it to*/
 input          Maternal_Name $&format..
;
run;

I cannot figure out a way to resolve the macro 'format' variable & I have
tons of segments to parse plus a large amount of records - therefore, I
only want to make one pass through the data.

Any suggestions???

Thanks!