If you want to seek names that begin with "WARD" you may use LIKE 'WARD%'
instead of using CONTAINS.

Bill Cook
Kent WA USA

----- Original Message -----
From: "Phil Nolette (NCS Group, Inc.)" <[EMAIL PROTECTED]>
To: "Rbase-L" <[EMAIL PROTECTED]>
Sent: Wednesday, October 10, 2001 8:05 PM
Subject: Code Share that Works for Name Extraction


> Most of the time I ask questions and never provide.  I know that many of
you
> are "Power RBase" developers and you might turn an eye on this code but,
it
> is working like a charm and I wanted to share it with you.
> What I had to do was to get the last name out of the full name and then
> check another table to find anyone with the last name using a CONTAINS
> statement.  The main problem was in the area of a name like "WARD".  I was
> getting all kinds of WARD (Hayward, Ward, Smithward, and ....). So I used
an
> SGET in the WHERE clause after I extracted the first character of the Last
> Name.  I now delimit to anyone with the Last Name containing Ward and
> nothing more.
>
>
> Phil
>
============================================================================
> ======
> -- ========Check NCA Data for Direct Dial Numbers Only  ======
>   COMPUTE vmany2 AS COUNT nca_id FROM usr WHERE  dir_dial_nbr = .vp4 AND
> mstr_rcd_typ = 'I'
>   IF vmany2 = 0 THEN
>     WRITE ' This CD has no individual match in NCA ten digit phone'
>     WRITE ' Checking for Name Equivalence on NCA '
>
>     -- Get the names related to this CD Record from NCA
>     -- Finds the position of the comma
>     SET VAR vsloc = (SLOC (.vp3, ','))
>     -- Increment the position by 2 to get first letter of first name
>     SET VAR vsloc = (.vsloc + 2)
>     -- Get the First Letter of the First Name
>     SET VAR vsget = (SGET(.vp3,1,.vsloc))
>     -- Extract the Last name from the full name
>     SET VAR vvncanm = (SSUB(.vp3,1))
>     -- Put it all together for the Search variable
>     SET VAR vsch = ((.vvncanm + ',' + ' ') + .vsget)
>     -- Get the first letter of the last name
>     SET VAR v1st = (SGET(.vp3,1,1))
>
>     DECLARE c5 SCROLL CURSOR FOR SELECT
> nca_id,comp_cc_cd,nca_nm,dir_dial_nbr,prim_ext_nbr +
>         FROM usr WHERE mstr_rcd_typ ='I' AND nca_nm CONTAINS .vsch and
> (SGET(nca_nm,1,1)) = .v1st
>
>     -- No Related Names Found in NCA
>     IF SQLCODE = 100 THEN
>       WRITE 'There were no associated names in NCA related to this CD
> Record'
>     ENDIF
>
>     -- Write the Related Records found in NCA for this CD Record
>     OPEN c5
>     FETCH NEXT FROM c5 INTO vnc51, vnc52, vnc53, vnc54,vc5primeext
>     WHILE SQLCODE <> 100 THEN
>       WRITE 'NCA=> ','ID: ',.vnc51,'COMP/CC =>',.vnc52,'Number =>
> ',.vnc54, 'NAME => ',.vnc53
>       FETCH NEXT FROM c5 INTO vnc51, vnc52, vnc53, vnc54,vc5primeext
>     ENDWHILE
>
>     CLOSE c5
>     DROP CURSOR c5
>
>
>
>   ELSE
>     WRITE 'Continue Analysis - Search NCA by CD ten digit '
>
>
>   ENDIF
>

Reply via email to