As an example, I have a User that has a list of Items that each have a name. I need to return all the users that contain the Item with a given name. I've seen ways to do this with annotations, but no examples using programmatic mapping. Is this possible?

Thanks,

Jeff

Of course it is possible. The programmatic mapping is as expressive as annotations. Seems you have a simple indexed embedded case:

Thanks...I realized it was in the documentation after digging a little further...oops. Although I do have a question. Given the model of a User with Items that have a string field "name", I mapped it like this:



Which works great as a query on User like this:



gives me all Users that have an Item with the name "SomeName". However, when I inspect the index via Luke, there are redundant fields. Both the "name" field and "items.name" field appear in the index (with the same number of terms, of course). I don't like this, because eventually I'll be indexing the item description which has a lot more terms and I don't want to have two sets like that. I want to keep the index as lean as possible. So, am I being redundant in my mapping or is this just the way things work? I should also point out that "items.id" shows up as a field in the index and I'll never need to use that. Any advice would be appreciated.

Regards,

Jeff

Hi,

that's because you create a second (independent) index for Item - mapping.entity(Item.class).indexed().indexName("item-index").
Try removing the highlighted part and just configure the properties.

--Hardy

Thanks Hardy, that worked like a charm.