HHH-19020 - provide more info in case a declared MappedSuperclass is not found #10739
Conversation
…not found Signed-off-by: Jan Schatteman <jschatte@redhat.com>
| @@ -274,6 +275,9 @@ public static void addPropertyToMappedSuperclass( | |||
| ClassDetails declaringClass, | |||
| MetadataBuildingContext context) { | |||
| final MappedSuperclass superclass = context.getMetadataCollector().getMappedSuperclass( declaringClass.toJavaClass() ); | |||
| if ( superclass == null ) { | |||
| throw new HibernateException( "Mapped superclass is null for declaring type " + declaringClass ); | |||
There was a problem hiding this comment.
This message is slightly better than an NPE, but it doesn't really tell the user what they did wrong, nor what they should do to fix it.
Also, shouldn't it be a MappingException?
There was a problem hiding this comment.
I think we might want to understand why the superclass is null, too, since looking at the code-path that leads here (from the Jira's stack trace) it looks like we're processing an embeddable that extends from a @MappedSuperclass. I fear that we might be processing the embeddable before its parent, I remember fixing something similar for https://hibernate.atlassian.net/browse/HHH-18103.
There was a problem hiding this comment.
I think we might want to understand why the
superclassis null, too, since looking at the code-path that leads here (from the Jira's stack trace) it looks like we're processing an embeddable that extends from a@MappedSuperclass. I fear that we might be processing the embeddable before its parent, I remember fixing something similar for https://hibernate.atlassian.net/browse/HHH-18103.
One of the issue's comments states "the superclass of the embedded id was not part of the mapped resources that Spring registers with Hibernate", so it seems like this was an error in Spring, and this was just an attempt to cater for the user's - I think correct - request for more useful (than an NPE) info on what happened.
@gavinking This is also why I chose to throw a more generic HibernateException, since we may not be sure what the cause of the missing MappedSuperclass is?
There was a problem hiding this comment.
I agree with @mbellade that we should dig a bit deeper into this, via the test case given here: https://github.com/lukaseckert/hibernate-test-case-HHH-19020/tree/main/orm.
One of the issue's comments states "the superclass of the embedded id was not part of the mapped resources that Spring registers with Hibernate"
I think you know what I think of the idea of taking at face value any claim made by a user in an issue report. :)
There was a problem hiding this comment.
My findings:
062afdb6cda9fe359740b0ed4aad2a1a3b83fc15 is the first bad commit
commit 062afdb6cda9fe359740b0ed4aad2a1a3b83fc15 (HEAD)
Author: Marco Belladelli <marcobladel@gmail.com>
Date: Thu May 23 11:46:59 2024 +0200
HHH-18103 Correct metamodel for embeddables with a mapped superclass
@mbellade thoughts?
There was a problem hiding this comment.
Before my fix, the metamodel for embeddables extending from mapped superclasses was simply wrong (the attributes were not considered as defined by the superclass).
That doesn't give me much context as to why the mapped superclass might be null, though. As I said in my previous comment, it would be nice to understand the cause of it as there might be a bug there - I can take a look at the issue and try isolating a test case if you're having trouble with that.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license
and can be relicensed under the terms of the LGPL v2.1 license in the future at the maintainers' discretion.
For more information on licensing, please check here.
https://hibernate.atlassian.net/browse/HHH-19020