Erick,
Thanks for the advice. I will take a look at
PerFieldAnalyzerWrapper to see if I want to take this on. For my
case, I have to use mexed case for a couple of fields since case
really does matter for them (ie apple is not the same as Apple), and I
actually don't want users to find the document if they enter the
"wrong" case (ie a search for Apple should not return docs containing
apple).
Bill
On 5/4/07, Erick Erickson <[EMAIL PROTECTED]> wrote:
Look at PerFieldAnalyzerWrapper. It allows you to use different
analyzers on different fields during the query parsing phase.
But I wouldn't go there if you don't have to. I suspect you'll spend a
LOT of time tracking down errors in your use of a mixed case index.
If for no other reason than your users will use the "wrong" case.
Unless your index is huge (and I don't consider, say, 8G huge), I'd
index everything in, say, lower case. And ditto for your query
parsing.
If you need to return data to the user in mixed case, then you can
*store* (but perhaps not *index*) the display fields. So you search
on one field and return data from another.
Best
Erick
On 5/4/07, Bill Au <[EMAIL PROTECTED]> wrote:
>
> I have an index with both fields that are case sensitive and
> insensitive. I
> am trying to use a QueryParser to accept query from end users for
> searching. The default behavior of QueryParser is to lowercase the prefix
> text to create the PrefixQuery. So wildcard search on the case sensitive
> fields does not work. If I use QueryParser.setLowercaseWildcardTerm
> (false),
> then wildcard search on the case insensitive fields does not work.
>
> Here is an example with two fields, name (case sensitive) and desc (case
> insensitive). The docment is
>
> name (case sensitive): PowerBook
> desc (case insensitive): professional mac laptop
>
> I want to be able to find the document with the following query:
>
> +name:Power* +field:Pro*
>
> Bill
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
