From: Sam Smoot <ssmoot@...>
Date: 2006-10-29T01:45:11+09:00
Subject: Re: Efficient parsing of large Excel documents in Ruby

I agree with Christian, but with a twist.

I like to use Excel Automation (w/ WIN32OLE) to convert the document to
a format that's easier to work with. That way, after the initial
conversion, you can handle CSV, XML, etc, and be working with a fast
native library instead of the slow interop.

Another reason I prefer this method is that Excel interop is sometimes
hard to get "just right". There's processes I've written with it
running for over a year that I've never had to touch, but there's
another that used to require I login to the server and kill orphaned
Excel processes. Someone changes a column, you don't code defensively
enough, and BAM!

So the simplicity of the Load/Convert process, and getting out of Excel
as quickly as possible, conserving resources and avoiding potential
bugs that might result in orphaned processes is a big advantage IMO.

Plus it's almost gauranteed to be the fastest, and simplest to develop.
(As long as you're OK with having to run on Windows and having Excel
installed anyways.)