Re: Problem with Case Sensitivity in Oracle?

From: Tony Kennedy <adk_at_scri.fsu.edu>
Date: 4 Mar 94 15:25:59
Message-ID: <ADK.94Mar4152559_at_ds4.scri.fsu.edu>

>> Paul Singleton <paul_at_cs.keele.ac.uk> writes:

   paul> There is a conflict between

   paul> * FANCY data: the way you want to store it for retrieval    paul> (e.g. nicely formatted, punctuated, in mixed case)

   paul> * CANONICAL data: the way you want to index that data (on its
   paul> essential informational content, ignoring layour,
   paul> punctuation, case etc.)

   paul> The best, and most general, solution is (surely?) to store

   paul> the information twice, in both canonical and fancy forms.    paul> You index the former, and retrieve the latter.

I think your idea can be generalized a little more. Define a hash code (say an integer value) which is guaranteed to be identical for all "equivalent" forms of the string, where "equivalent" is application-specific, and index the data by hash code. Received on Fri Mar 04 1994 - 15:25:59 CET