[thesite] just wondering..

Seth Bienek seth at sethbienek.com
Wed Feb 28 08:15:23 CST 2001


Hey Dan,

Someone pointed me to that when I first got started on the project.

The custom tag I wrote parses rfc822-formatted text and returns the results
either as a query recordset or an XML file.

It follows the design document guidelines as far as getting the data into a
database; pulling the data out and displaying it in different ways is the
easy part. :)

As far as I know, there are currently two hold-ups on the project.  The
first being that there is not a database to put the tips into;  well, there
IS a database, but it doesn't match the data that we want to pull out of the
lists, or something.  I haven't worked with it in awhile.  The second
hold-up is the performance hit that the webserver takes parsing through the
megabytes upon megabytes of text in the .mbox files, and more importantly,
the database inserts.  The tag parses about 8 emails per second on my
p3-500, 256Ram NT box; it performs similarly on l.e.o., but when we added
the database insert, it slowed to about 10 seconds per tip.

I could run the process on my machine at home on downloaded mbox files, and
then upload an xml file with the results (if you have a way to get the XML
into Oracle), then run daily's on l.e.o. to process new messages, to keep
from tying up l.e.o., but I'm still concerned about the time it's taking to
get the data into Oracle.

I had thought about rewriting the parsing function in PL/SQL, but I don't
know PL/SQL well enough, and don't have the time to tinker with it.

I'm open to suggestions.. :)  Do you still have that XML file I sent you?

----- Original Message -----
From: Daniel J. Cody <djc at starkmedia.com>
Subject: [thesite] just wondering..


> seth, if you ever saw this:
> http://www.easyweb.co.uk/thetips/reqs.html






More information about the thesite mailing list