Hi,

I'm using Hibernate3 for persisting POJOs. Each object in the object graph has a surrogate key which is generated using one of the Hibernate ID generation policy. I'm frequently required to update complete object graphs recreated from a different source(and hence it does not have its surrogate(s)) to the DB.

I'm wondering what is the best way to updated such an object graph -

1. Should I fetch the complete object graph from the database and update every property of every persistent object in the graph with values from the non persistent object graph(the one with updated values). This approach would be cumbersome when it comes to handling complex graphs.

2. Fetch respective IDs (using SQL) of object graph and assign them to the non persistent object graph.

Let me know if you have a better approach.

Thanks in advance.
AYB

You might want to consider doing something tricky with Apache Commons' Bean Utils to help out, but there is no easy way to do this if your objects are persistent but do not have ids. It may also be worthwhile to look into a way to pass the id values between you and the source of the changes.

--
Tim

It is not possible for the generator of the updated object graph to pass the surrogate ids as it not aware of them. Also the generator will _always_ generate full object graph in case of both insert and update.

Could you plz elaborate on Common Bean Utils approach you suggested.

Thanks,
AYB

Had a thought, a mix of property naming conventions and reflection can help updating such object graphs in a generic and concise way.