Hello Hibernators

I have two tables, VENDOR_SIZE_CHART and VENDOR_SIZE_VALUE. Each VENDOR_SIZE_CHART is associated with many VENDOR_SIZE_VALUES.

VENDOR_SIZE_VALUE has a SORT_ORDER column, I would like Hibernate to use this column to order the collection every time its retrieved. This column can have gaps (i.e. 1,2,3,<gap>,5,6,7) and duplicates. I don't really care about tie breaking order in case of duplicates, but there is a primary key ID column to use.

The end user can update SORT_ORDER through the web application, if that happens the next time VENDOR_SIZE_CHART is updated the VALUE's should be retrieved in the new order.

I'm having trouble implementing this in HBM. Here is my mapping:


I'm having the following problems:
- I get an ArrayIndexOutOfBounds error when there are duplicate sort order values
- null elements are being placed in the list where the gaps occur, as if the collection was an fixed length array indexed by sort_order

Is there a way to tell Hibernate to not place null elements in the list, and to use the primary key as a tie breaker in case of duplicates?

Thanks!

Try with a <bag> instead of <list> and use the order-by attribute for the sorting. Something like:

Thanks! What should the type of the collection be in the POJO? Should it be a List or a SortedSet?

A <bag> gives you a list. I guess a <set> mapping could be used instead if you prefer that. If you use the <set order-by...> you'll get a LinkedHashSet. You can also use <set ... sort="..."> where sort is the class name of a Comparator implementation. In the latter cse, the sorting is done in memory and gives you a SortedSet. See http://docs.jboss.org/hibernate/stable/ ... ons-sorted for some more examples.