We currently have a solution for this in production, but it's pretty bad. I'd like a fresh perspective on it please.

We have a PERSON table with an ID, LAST_NAME, FIRST_NAME, MIDDLE_NAME, and SUFFIX_NAME. All these columns (except ID of course) are nullable.

What is the best way to index this table, and what is the best way to query for full name with wildcards?

This is used in a web app, and we need to allow the user to search for something like

Smit*, Bob

and it return "Bob Smith", "Bobby Smithson", etc.

We will not allow the user to sort results.


You have two options I can think of:

  1. Index for a specific set of user searches


  2. Full Text Search

For option 1, you need to convert your search string "Smit*, Bob" into something you are indexing on, eg: where upper(last_name) like 'SMIT%' and upper(first_name) like 'BOB%' with function based indexes on upper(last_name) and perhaps upper(first_name)

| improve this answer | |
  • +1 Here is the conversion: Replace(UPPER(:FirstName), '*', '%'). – Leigh Riffel Aug 26 '11 at 14:32

We use Lucene.Net to search for contacts on our Oracle database. It's a bit of an investment in coding to set up the indexer, index and searched fields but once it's done you take advantage of the many search options.
Users report the usual problems: enter too wide a search term and get back too many results but it is really good for names where the name entered is, for example: "Robertson" but your search is "Bert"

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.