Path: utzoo!mnetor!uunet!lll-winken!lll-tis!ames!pasteur!ucbvax!ucdavis!iris!ambrose
From: ambrose@iris.ucdavis.edu (Paul Ambrose)
Newsgroups: comp.databases
Subject: Re: dBaseIII+
Message-ID: <1785@ucdavis.ucdavis.edu>
Date: 19 Apr 88 17:18:41 GMT
References: <2413@mandrill.CWRU.Edu> <640@naucse.UUCP>
Sender: uucp@ucdavis.ucdavis.edu
Reply-To: ambrose@iris.UUCP (Paul Ambrose)
Organization: U.C. Davis - College of Engineering
Lines: 35
Keywords: dBaseIII+, Existential Quantifiers

In article <640@naucse.UUCP> wew@naucse.UUCP (Bill Wilson) writes:
>In article <2413@mandrill.CWRU.Edu>, ananth@mandrill.CWRU.Edu (ananth srinivasan) writes:
>> I have a problem with dBaseIII+ syntax and would appreciate any help.
>> I have two relations with join-able attributes (say ID) and I need
>> to find out those ID's that exist in one and don't exist in the other.
>> In SQL I can use the NOT EXISTS syntax; in QUEL I can use the
>> ANY...=0 syntax; and in Rbase V I can use the SUBTRACT operator
>> (which really makes the most sense). My question is what is the
>> appropriate syntax in dBase?

My DB3 is rusty, but I know the following works:

Given two tables, A and B, you can determine records in A that do not
have corresponding IDs in B by:

	select 1
	use B
	index on ID field
	select 2
	use A
	*(Index A here if desired)
	set relation on ID into B 
	list field_list_of_A for id <> B->id

The *for* condition (<>) will be true for all those 
records in A that do not have a correspponding ID's in B.  

Obviously, reversing the A's and B's will give the opposite, 
that is, all records in B and no corresponding IDs in A.

In both cases, the statement:
	list field_list_of_A for id = B->id
will give those records with IDs in both tables.

Of course, you can use count, report, label, etc., 
instead of the list statement.