Re: Intermedia question

Couple of things:

1. You must be running the context engine as a daemon, otherwise, the indexes are not updated. RTFM :-). You can stagger the rebuilding of your Context indexes if you wish to reduce the impact of the intermedia engine.
2. You must set the filtering parameter on the Context index itself to read the MS Word file. If you do not filter the file, it will not be searchable, or other data will end up as keywords

CREATE INDEX resume_content ON resume(CV) INDEXTYPE IS CTXSYS.CONTEXT PARAMETERS('filter ctxsys.inso_filter');

3. You should learn how to set storage preferences for the Intermedia text engine, or the Context indexes ***WILL*** become a bottleneck.

4. You should learn how to set custom lexer preferences so that the keyword indexes will properly index strings such as C++, PL/SQL, J++, etc. Otherwise, these strings are considered identical to C, PL, SQL, J.

Chris Weiss
Chief Scientist for Database Engineering PureCarbon, Inc.

<thebat88_at_hotmail.com> wrote in message news:92h558$k8i$1_at_nnrp1.deja.com...
> Let say I create a table RESUME(id number, CV BFILE)
> with id as primary key and create an index on this table
> as create index my_index on RESUME (CV) indextype is ctxsys.context;
>
> I insert a row into the table as
> INSERT INTO RESUME(1, bfilename('docdir', 'my_resume.doc');
> where docdir is a mapping to the directory where the resume exists.
>
> When I do a query to see if the resume CONTAINS a word (e.g. Java) it
> doesn't return anything even though the word document does contain the
> word Java.
> SELECT id from RESUME WHERE CONTAINS(CV, 'Java') > 0;
>
> Any help will be very much appreciated.
>
>
> Sent via Deja.com
> http://www.deja.com/
Received on Thu Jan 04 2001 - 10:15:50 CST