[thelist] XML based site search

Nik Schramm n at industriality.com
Wed May 8 16:46:01 CDT 2002

Thanks for the helpful comments Jonathon and Avi. You are probably right
in suggesting an index-based search system to spare the server any
unnecessary processor load, especially as that is not that hard to
implement automatically. If the site were to get *very* popular in the
medium run, then I agree that an outside search engine would probably
seem like a good idea.

FWIW (and because it interested me) I put together a quick-'n'-dirty
version of the XML based search method I outlined in my original post.
Basically, it loads a series of XML docs and uses the .NET XML InnerText
property to dump the text content of said Doc into a string, which I
then compare to my search term using IndexOf(). Amazingly this is very
fast, less than one second for 23 files.

Because time is an issue here, I'll probably go with that for the time
being, while at the same time preparing myself for the need to create an
index-based and/or external variation along the lines of what you both



www.industriality.com - candy for the inner eye

More information about the thelist mailing list