OpenEuphoria

1. RE: dAtabase

• Posted by Matthew Lewis <matthewwalkerlewis at YAHOO.COM> Jan 11, 2002
• 405 views

> -----Original Message-----
> From: Aku [mailto:aku at inbox.as]

> What database program is suitable for euphoria (simple) but
> I dont like EDS because no indexing support. Just database
> for local files, not database with server etc.

Aku,

You could check out EuSQL in the archives.  It sits on top of EDS, providing
a SQL (Structured Query Language) front end, turning EDS into a relational
database.  Much of it is still pretty rough, and the documentation isn't as
clear on some things as it should be, but I'm happy to respond to any
questions anyone might have.  It currently does not support indexing,
although it should shortly.  I'll be picking it back up here soon to work on
it.  I'd be interested to hear more on what you plan to do, and why you're
interested in an indexed database.

Here's a little bit on SQL, in case you're not familiar with it.  First,
it's probably the most common database language, since it's used by Oracle,
MS SQL Server, Informix and DB2, to name some of the biggies.  It allows you
to manipulate data in an almost english like manner:

"SELECT NAME FROM EMPLOYEES WHERE NAME LIKE 'A*';"

This command would return a list of employee names starting with 'A'.  There
are similar ways to insert, update and delete records.  The real benefits
come when you exploit the relational aspect of the database.  It allows you
to establish relationships between tables.  This means that you can store
the same amount of information in less space, but still have an easy way to
retrieve it.

Suppose you have a table of employees, containing (but not limited to)
names, and another containing all of your company locations.  You could have
put the location information in the employee table to keep track of where
everyone works, but this  could lead to lots of errors (not to mention a lot
of duplicated data).  But if you give each location an id (a number is nice,
since it doesn't take up a lot of space), and put the id in the employee
table, you save a lot of space.  We do this all the time with linked
sequences, but it's a bit more complex with a database.  Here's how you
might get a listing of employees and their locations:

"SELECT EMPLOYEES.NAME, LOCATION.CITY, LOCATION.STATE FROM EMPLOYEES INNER
JOIN LOCATION ON EMPLOYEES.LOC_ID = LOCATION.LOC_ID"

While EuSQL doesn't currently support indexing, it can make data
retrieval/manipulation much easier through queries.  Also, it allows you to
structure tables into fields and subfields.  The EDS key is used as the
EuSQL primary key, which uniquely identifies each record.  As a user of
EuSQL, you don't have to worry about the difference between key/data as when
you're directly using EDS.

I think it makes a nice compromise between Euphoria's flexible sequences and
a database's structured fields.  You can embed fields within fields
(atoms/sequences within sequences) and access them using dot notation:

NAME.FIRST

Currently, no data type checking is done, although that's already
implemented for the next release.  Database creation is manual right now,
although this will be automatically handled in the next release.

Matt Lewis