Re: Like and Case Sensitivity

I'm not so sure all systems can turn case sensitivity on or off easily.

Some do other things than simply case sensitive or case insensitive by
default.  By default Verity is case sensitive if the query term has
mixed case and insensitive otherwise, but does have a case
operator.

Some other full text systems don't necessarily have a case
operator so they wouldn't easily be able to conform to the
standard if there is no out for being only case sensitive or  case
insensitive..

Do we need to include query discovery for control of case
sensitivity capabilities?

--Rick

Kevin Wiggen wrote:

> <case sensitive attribute on like> it's missing.  It's probably an
> oversight.  There is no reason not to allow it.
>
> Note that the default for all other casesensitive attributes is 1.
>
> 1) should the default for like be the same as the usual default?
>
> <KW> I see no reason why the default should be different.  It should be the
> same for consistency if for nothing else
> </KW>
>
> 2) should we change the usual default from 1 to 0?  (that is, comparisons
> are case insensitive unless explicitly stated otherwise?)
>
> <KW> Most systems I know of have an easier time doing case sensitive
> searches (or its less CPU cycles at least), I think we should leave the
> default where it is.  This way the default is less CPU intensive.  Its easy
> to override so it should not be a big deal
> </KW>
>
> Kevin

Received on Wednesday, 2 February 2000 13:27:44 UTC