Hi David, I don't think the extra column will cause a problem for SynchronizeMappings, unless you tell it to recreate the database each time. If you just have it build the schema it won't recreate the table/
Something like this should work for you : <property name="openjpa.jdbc.SynchronizeMappings" value="buildSchema" /> -Mike On Mon, Mar 31, 2008 at 8:08 AM, Adam Hardy <[EMAIL PROTECTED]> wrote: > David Goodenough on 31/03/08 13:37, wrote: > > On Monday 31 March 2008, Adam Hardy wrote: > >> David Goodenough on 31/03/08 11:06, wrote: > >>> I have an application for which I would like to use OpenJPA, and it > >>> involves the PostgreSQL tsearch(full text search) facility. > >>> > >>> Now obviously I would not expect OpenJPA to understand or use > (directly) > >>> tsearch, so I assume that I can do the actual searches using the > facility > >>> that OpenJPA has to use real SQL. > >>> > >>> But in order to use tsearch, I need to set up a column of type > tsvector, > >>> which is course is not something that OpenJPA can be expected to > >>> understand. So the question is how do I tell it to ignore it, i.e. to > >>> allow it to exist but not to try to use it. > >> Co-incidentally I came across a problem with an unmapped property on an > >> entity causing OpenJPA to grind to a halt, until I marked it as > transient, > >> but I think this is the other side of the coin compared to your > situation. > >> > >> You don't actually say what you do or do not want to do with the > column. > >> > >> Presumably you don't have a property on your java entity for the column > and > >> you are not mapping it? > >> > >> AFAIK JPA will ignore columns that are not mapped, so you should be OK, > but > >> I guess you're asking because it's not OK? > > > > Actually that is exactly what I want, I just could not find the place in > the > > manual that said it. > > > > I need to ignore the column because I do not need to try to map a column > > of type tsvector into a Java type, as I will only use it in real SQL > searches. > > The results of these searches only need the "normal" columns. > > > > That is fine for normal operations, but is there any way to mark a > column as > > to be retained if I use openjpa.jdbc.SynchronizeMappings? I would not > want > > to do this normally, just as part of an upgrade process so that the DB > is > > brought from whatever state it is in to the current schema. > > Sorry, never used openjpa.jdbc.SynchronizeMappings. Don't know. Maybe > someone > else here does.... >
