Question

Our customer’s data (SQL Server 2005) has html entities in it (é -> é).

We need to search inside those fields, so a search for “équipe” will find “équipe”.

We can’t change the data, because our customer’s customers can edit those fields as will (with a HTML editor), so if we remove the entities, on the next edit they might reappear, and the problem will still be there.
We can’t use a .net server-side function, because we need to find the rows before they are returned to the server.

I would use a function that replaces the entities by their UTF-8 counterparts, but it’s kind of tiresome, and I think it seriously drops the search performances (something about full table scan if I recall correctly).

Any idea ?

Thanks

Answer 1

You would only need to examine and encode the incoming search term.
If you convert “équipe” to “équipe” and use that in your WHERE/FTS clause then any index on that field could still be used, if the optimizer deems it appropriate.