Subj : Re: Database program to just work with text files?
To   : comp.os.linux
From : Average_Joe
Date : Mon Sep 06 2004 09:26 pm

In article <6f1027e6.0408210953.2bab0d89@posting.google.com>, Adam wrote:
> Hello,
> 
> Does anyone know if there exists a database program for Linux that
> just modifies one or more text files?

bash/cut/awk/sed/grep  ? :-)

Seriously, have a look at:

http://www.cse.ucsc.edu/research/compbio/rdb/

The idea is to use UNIX pipes and tab files in a RDB. I used it many
years ago and liked it because of it's simplicity.

It's not SQL, it's a totally different way of handling data, that may be
a plus or a minus to you. (I kind of liked it)

I've done many similiar things myself in perl for application specific
data.
 
> I've been using mysql, but I don't want something for networks; I
> don't like the daemon that runs in the background (according to top
> and KDE System Guard, it has 4 threads running taking about 25MB
> each!).  I just have a personal "database" consisting of tab-separated
> values and I want to query it with SQL statements rather than look
> through the file.
> 
> All I want is a program to do the following:
> 1. open 1 or more text files, 1 file per database table
> 2. read the data into memory
> 3. execute an SQL statement on it
> 4. if the statement was a modification, save the data back
> 
> Permissions would be handled by the file permissions and owner(s), no
> daemons, no networks, no sockets, just nice and simple.

Except for the SQL requirement, RDB is nice for what you describe.

There was once a project that used _nothing_ but shell scripts to
implement an SQL database. Maybe someone will remember it. ('course SQL
via shell is more or less just a way of proving it *can* be done,
doesn't mean it *should* be done :-) )

Jamie
-- 
http://www.geniegate.com                     Custom web programming
User Management Solutions                   Perl / PHP / Java / UNIX

.