Regression: Constraint Validations are no longer translated into DataIntegrityViolationException with neo4j-java-driver >= 4.3.2

[GrahamLea]

Scenario

Execute a modifying Neo4j transaction that results in a constraint violation.

Expected behaviour

The repository throws a org.springframework.dao.DataIntegrityViolationException.
(This is what it does with neo4j-java-driver versions < 4.3.2)

Actual Behaviour

The repository throws a org.neo4j.driver.exceptions.ClientException.

Reproduction

1. Clone https://github.com/GrahamLea/neo4j-bugs
2. Load in IntelliJ IDEA
3. Run TestNodeTest.raises DataIntegrityViolationException on constraint violation
4. It will fail
5. In pom.xml, change neo4j.driver.version back to 4.3.1
6. Run test again and it will pass

Comments

It's possible there's a regression in neo4j-java-driver that has caused this to happen.
But the expected result DataIntegrityViolationException is part of this project, so I didn't feel I could submit the issue there.

[michael-simons]

Puh, this is on a very old SDN version that we actually don't support anymore. There's actually little I can do, VMWare most likely won't do any release of the old 5.x branches anymore.

However, I think this can be fixed in the underlying OGM.

OGM 3.2.34 will go out this morning, please try it out and close the issue if this fixes things for you.

[GrahamLea]

Thanks for jumping on that so quickly.
Should I expect that 3.2.34 release to pop up in Maven Central (it hasn't yet), or do I need to pull it from somewhere else?

[michael-simons]

That's kinda odd, because our CI says the job has finished correctly and normally, central isn't that slow.
In short: Yes, it should have been on central.

[michael-simons]

Ok, it was a problem of the operator (me, in this case) which is solved now.

https://repo1.maven.org/maven2/org/neo4j/neo4j-ogm-core/3.2.34/

Thanks for your ping, @GrahamLea

[GrahamLea]

Yep, 3.2.34 has fixed it. Thanks. Amazing service! 😉👏

[michael-simons]

Thanks a lot for your feedback <3