Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!pacific.mps.ohio-state.edu!linac!midway!ellis.uchicago.edu!goer
From: goer@ellis.uchicago.edu (Richard L. Goerwitz)
Newsgroups: comp.lang.icon
Subject: Re: Records
Message-ID: <1991May31.143815.4204@midway.uchicago.edu>
Date: 31 May 91 14:38:15 GMT
References: <31.May.91..11: 23:12.bst..060024@EMAS-A>
Sender: news@midway.uchicago.edu (NewsMistress)
Distribution: inet
Organization: University of Chicago
Lines: 22

In <31.May.91..11:23:12.bst..060024@EMAS-A> R.J.Hare@edinburgh.ac.uk writes:
> 
>Next question - I am using records as the basis for a simple database
>program. Given that the program will be written in a tight enough way to
>ensure that it won't be possible to write a list with more elements than
>there are in a database entry, what is the advantage of using a record
>to store a database entry as opposed to a list?

Record field access is much faster than list subscripting.  You can also
get at the fields by name, and not worry about their order.  If you want,
though, you can subscript records (as long as you realize that doing so
erases most of their method of/speed of access advantages).

If you really want to be clever, use a record for all the fields you know
you'll be using, but make one member of the record into a list where you
can store optional or other fields.  You'll get the best of both worlds,
at the expense of some clarity and straightforwardness.

-- 

   -Richard L. Goerwitz              goer%sophist@uchicago.bitnet
   goer@sophist.uchicago.edu         rutgers!oddjob!gide!sophist!goer