Re: Using contrib/fulltext on multiple tables. - Mailing list pgsql-bugs

From Eric Johnson
Subject Re: Using contrib/fulltext on multiple tables.
Date
Msg-id 015301c35b04$62db4580$1a08090a@MINIDEV
Whole thread Raw
In response to Using contrib/fulltext on multiple tables.  ("Eric Johnson" <ej@ejinnovations.com>)
List pgsql-bugs
I just noticed something about how this is behaving.  When I run the script
on my machines, it's putting the index data tb_c_ingredient into
tb_c_step_fti.  If I exit psql, go back in and manually insert, it goes to
the correct table tb_c_ingredient_fti.

I wrote a much simpler script (see attachment: just  3 tables being indexed,
3 triggers, and 3 index-data tables) and am having the same problem - only
all the data is going into the first table's index until I logout and go
back in.

It may be worth noting the following error if I run the script multiple
times in the same psql session:

psql:bug.sql:66: ERROR:  pg_class_aclcheck: relation 125816 not found
psql:bug.sql:67: ERROR:  pg_class_aclcheck: relation 125816 not found
psql:bug.sql:68: ERROR:  pg_class_aclcheck: relation 125816 not found

That's happening on the INSERT INTO table1 ...    table2 ... and table3
statements.


Eric

----- Original Message -----
From: "Tom Lane" <tgl@sss.pgh.pa.us>
To: "Eric Johnson" <ej@ejinnovations.com>
Cc: "Pgsql-Bugs" <pgsql-bugs@postgresql.org>
Sent: Sunday, August 03, 2003 8:29 PM
Subject: Re: [BUGS] Using contrib/fulltext on multiple tables.


> "Eric Johnson" <ej@ejinnovations.com> writes:
> > Later in the script when inserting into these tables via stored
procedures,
> > the full text indexes get created for a and b but not c.  It's just
empty.
>
> I can't reproduce this in either 7.3 or CVS tip.  The index tables seem
> to have reasonable numbers of entries in them after running your script:
>
> recipe=# select count(*) from tb_c_step_fti;
>  count
> -------
>     41
> (1 row)
>
> recipe=# select count(*) from tb_c_recipe_fti;
>  count
> -------
>    207
> (1 row)
>
> recipe=# select count(*) from tb_c_ingredient_fti;
>  count
> -------
>    103
> (1 row)
>
> If that wasn't what you meant, you'll need to be more specific.
>
> regards, tom lane
>

pgsql-bugs by date:

Previous
From: Tom Lane
Date:
Subject: Re: html tarball not found in doc/Makefile
Next
From: Bruce Momjian
Date:
Subject: Re: small bug in op + between datetime and integer