Using projections for persisting data

[gonzalad]

This issue is related to a the following comment where we mention using projections for persisting data.

Is this feature already planned ? And if yes, in what milestone ?

#2305 (comment)

small preview of the future: There will probably nested projections coming up.
So you can describe your graph model "completely" in SDN but for load and persist operations
use projections as a kind of blueprint for the data that you really want to fetch or persist.

This will improve not only the performance of fetching data (by just not fetching unneeded data)
but also the way data gets persisted without touching / updating unchanged parts.

I have this question because I'm migrating an application from SDN5 to SDN6.

The application has a deeply nested structure (tree-like entities)
In SDN5 I could persist the entities without having to load all the entity graph.
In SDN6, this is not possible anymore (so even changing a simple attribute in a node would require me to load all the graph).

So the idea for me would be to:

• keep my current tree-like entities (to use when I need to manipulate the whole tree).
• have projections for node manipulations, and use those projection for queries but also for persistance operations

Thanks !

[meistermeier]

Yes, this is already available since 6.1.3 (https://github.com/spring-projects/spring-data-neo4j/releases/tag/6.1.3)
Some documentation on the topic:
https://docs.spring.io/spring-data/neo4j/docs/current/reference/html/#projections.sdn.multi-level
https://docs.spring.io/spring-data/neo4j/docs/current/reference/html/#projections.sdn.persistence

If this solves your issue, please close it. If you have any further questions, you can reach out on StackOverflow or our Discourse https://community.neo4j.com/c/drivers-stacks/spring-data-neo4j-ogm/30

[gonzalad]

Thanks very much @meistermeier, I don't know how I managed to miss it in the docs !