One of the blogs I read by Dominic Delmolini had an interesting question that seemed an nice exercise in the use of sql and probably analytic sql. It turns out that an elegant solution (by the questioner himself) uses the 10g addition of reg_exp_replace as well. I have been using regular expressions a lot in unix scripting and have grown fond of them, although they are terrible to read if you haven’t made them yourself.
The solution presented by Dominic made me think about uses for a query of this kind in the current day and age. Who still has a need for an encyclopedia after all, these days? Other than that it looks so nice on your bookshelves?
The thing that I came up with would be something along the lines of a materialized view for the contents of, for example, an enormous parts database with millions of parts, bolt and nut like parts that is. With this query (or materialized view based on it) you could present someone who is looking for a part but does not know what it is called exactly, with a quick way to find what he/she is looking for. You could also use this in a datawarehouse to present an overview of the number of people selected for a mailing of some kind per part of the city – or street names the selected people live in. I think that would be a rather intuitive way of showing results that many people are used to, but, up to now, was not so easy to create.
Another thing learned.
Recent Comments