Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.misc -> Re: Indexing for Oracle Text
Hi Can,
When you are building the text index oracle is storing the tokens in DR$<>$I and they are store either in UPPERCASE or MIXEDCASE (lexer pref: MIXED_CASE). If you have your index created with MIXED_CASE=YES then both queries will find the correct documents:
because oracle is using UPPER(tokens) and then search in the index.
Is that what you wanted to know ?
Thomas
Can Oezdemir wrote:
> Hello,
>
> I need to index a column of type BLOB, containing html text data as well as
> pdf and word documents. The indexing wasn't the problem, but now I'm trying
> to find a way how to get Oracle to search without considering the case of
> the the text (case-insensitive).
> Unfortunately I can't just say: ...WHERE CONTAINS(UPPER(columnname),
> 'searchString'.....
> because the upper function doesnt' support blob, and besides, it wouldn't
> work anyway, for some of the data is saved in pdf format...
>
> Thus, I guess the only possibility to do a case-insensitive search would be
> to index the column in lower case, right? As for now, I'm using the Oracle
> INSO Filter. But how can I additionally filter the columns (*after* those
> having run through the INSO filter), so that the index merely contains lower
> case text?
> I'm thinking of Oracle's procedure filter, or user filter, but I don't
> exactly know how one can apply multiple filters (INSO + lower-case
> filter)... - does anybody know how to accomplish this? Any ideas would be
> appreciated :)
>
> Thanks in advance,
> Can
>
Received on Thu May 29 2003 - 16:24:11 CDT