home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
icon
/
newsgrp
/
group94a.txt
/
000169_icon-group-sender _Sat Jun 11 01:15:37 1994.msg
< prev
next >
Wrap
Internet Message Format
|
1994-08-19
|
2KB
Received: by cheltenham.cs.arizona.edu; Sun, 12 Jun 1994 14:14:29 MST
Via: uk.ac.edinburgh.festival; Sat, 11 Jun 1994 09:15:04 +0100
Date: 11 Jun 94 09:14:58 BST
From: R J Hare <rjhare@festival.ed.ac.uk>
Subject: Databases
To: icon-group@cs.arizona.edu
Organisation: Edinburgh University Computing Service
Message-Id: <9406110914.aa19146@uk.ac.ed.festival>
Status: RO
Errors-To: icon-group-errors@cs.arizona.edu
Thanks to all who reponded to my query about databases.
Suggestions covered a wide range of options and included:
Use a flat file database
Use Shoebox (or some other proprietary db)
Use BiBTeX (I actually already use Scribe for my works db of papers)
Use Awk (example in the Awk book)
Use the Icon db progams in ipl
Use the example in Alan Corres' book
I also had a look at:
Programming Tools (K & R)
Algorithms in Snobol 4 (Gimpel)
String & List Processing in Snobol 4 (Griswold)
In the end, for various reasons to do with portability, simplicity, cost,
compute-literacy of the end-users (well, some of them - probably including
me!), and, finally, the fact that I like programming in Icon, I will be going
for a combination of the first and last options, ie: a flat file database
using the db program in Corres' book as a starting point.
This means that I can store the database as a simple text file with lines of
the form:
author|initials|title|publisher|date|isbn|price|date_of_entry
and read such records into whatever Icon structure is necessary for the task
in hand. For example to add records, I don't even need to read the db in, I
just open the file in "a" mode. When I want to sort the file, I just call in
the system sort. I only really need to read in the whole shootn' match when I
want to do a search or report.
This approach will get me on the air asap, but allows for more sophisticated
programming as I have new ideas - all that is necessary is that I maintain the
format of the flat db file (which I can't see any reason to change - all the
information I want about my book collection is there).
I haven't yet given much thought to how to modify this to cope with
discs/tapes/cds which have pieces by different composers on them.
If anyone is interested, let me know and I will keep them informed about this
(and I stress) simple project.
Roger Hare.