🚀 go-pugleaf

To create an "on-the-fly" link to two or more different databases,
the most elegant solution I know of is based on the use of pointers
and arrays.

Note, the following is being typed in from memory so minor syntax errors
may creep in...

Firstly, let's assume you have files [zip] and [addr1] or [addr2] used
as source for addresses, each with fields for Name, Address, City & Zip
(they DO NOT have to have the same names or be in the same position for
this technique to work)

Note: I'm going to use >> to indicate the 4D pointer (option-\) char.
                                               
Define an array to point to the relevant fields
ARRAY POINTER (addrFlds; 2; 4)
addrFlds{1}{1} = >>[addr1]Name
addrFlds{1}{2} = >>[addr1]Address

and so on...
addrFlds{2}{4} := >>[addr2]Zip


our [Zip] database has an integer field used to indicate which address
file is relevant (ie: it always contains 1 or 2).

To reference a given address, simply use
addrFlds{[Zip]whichAddr}{1}>>       
addrFlds{[Zip]whichAddr}{2}>>       
addrFlds{[Zip]whichAddr}{3}>>       
addrFlds{[Zip]whichAddr}{4}>>       

This technique has lots of other applications - I've just finished a 
system for converting ArchiCad output files to a jobcosting file, which
is totally user-defined. They can define "factors" that describe how to
extract data, ask the user or perform a calculation. Using the above      
technique I have one master [Factors] file and a number of "variant"
files. This is very much like object-oriented programming in that you
can add new Factor types without changing a bit of existing code!

If anyone's interested in more detail, I'll post again, enough dribbling
for now!

Andy Dent (A.D. Software - Mac & VAX programming)
94 Bermuda Dve, BALLAJURA  Western Australia  6066
Phone/Fax: 09 249 2719 (local)  +619 249 2719 (International)
       Internet: dent@DIALix.oz.au    Compuserve: 100033,3241