[Spring Data Graph] Separate full text search indices for different parts of graph?

[loki2302]

Consider this code:

@EntityNode
class Document {...}; // just a text

@EntityNode
class User {...}; // every user has its own documents and its friends

User user1 = userRepository.findByPropertyValue("name", "user1");
Iterable<Document> docs1 = documentsRepository.findSingleUserDocumentsByTitle(user1, "title here"); // only considering user1's documents
Iterable<Document> docs2 = documentsRepository.findOwnAndFriendsDocumentsByTitle(user1, "title here"); // user1 itself and its friends
Iterable<Document> docs3 = documentsRepository.findAllDocuments("title here"); // looking EVERYWHERE

What's the right approach for docs1 and docs2? The third one, I believe, is just about having @Indexed with full text search flag, but what about others? In other words, is it possible to create indices based on objects (sets of objects) and not on classes?

[MichaelHunger]

Right now you can either use the default index names which are the class-name or provide an index name in the @Indexed annotation.

To use an index derived from properties of a set of instances, you'd have to either index the entities manually or override the IndexingPropertyFieldAccessorListener which takes care of indexing (or register your own).

Right now it determines the index statically on the field but it can be changed so that this choice is made at runtime depending on properties of the entity.

If you need any more help just ping us again.

Cheers,

Michael