allow null value persistence and retrieval #111

NathanQingyangXu · 2025-07-11T20:22:53Z

https://jira.mongodb.org/browse/HIBERNATE-48

The ticket has both tech design part and implementation part:

Tech Design doc: http://go/hibernate-48
This PR is the implementation part

The main decisions from design are summarized as below:

treat null and missing fields interchangeably
no workaround if we are sure some defact is from Hibernate ORM side; try to report and let Hibernatem team fix it instead

The main work in this PR includes:

remove all previous various exceptions throwing with HIBERNATE-48 reference
enrich existing testing cases by inserting null values (particularly to the sorting and array/collection integration testing classes; comparison operator testing classes won't be touched for it belongs to the scope of HIBERNATE-74 or SQL nullness ternary logic emulation)
some defacts or tech improvements

Some tech decisions in this PR:

only test explicit null field for tech design has finished the validation that missing fields and null fields are interchangeable for find expression
enabled the support of @DynamicInsert for two reasons: 1) it was disabled for protective development reason during previous PR; now the reason ceases to exist; 2) we can't avoid mixture of null and missing fields even if we disabled @DynamicInsert for HQL partial insertion statement will end up with the mixture easily

NathanQingyangXu · 2025-07-11T20:27:12Z

src/main/java/com/mongodb/hibernate/jdbc/MongoPreparedStatement.java

+
+    private static final Set<String> COMPARISON_OPERATOR_NAMES_SUPPORTED =
+            Set.of("$eq", "$ne", "$gt", "$gte", "$lt", "$lte");
+


HIBERNATE-74 exceptions are intertwinced with HIBERNATE-48; we have to refrain from throwing HIBERNATE-74 exceptions from JDBC flow to unblock HIBERNATE-48. For this reason, I decoupled them and created a separate HIBERNATE-74 protected mechanism as above. It is more robust and is free of cache implication.

Given HIBERNATE-74 is to be done soon, we might well simply ignore the above protection.

NathanQingyangXu · 2025-07-11T20:30:08Z

src/main/java/com/mongodb/hibernate/internal/type/MongoStructJdbcType.java

        if (domainValue == null) {
-            throw new FeatureNotSupportedException(
-                    "TODO-HIBERNATE-48 https://jira.mongodb.org/browse/HIBERNATE-48 return toBsonValue(domainValue)");
+            return null;


we have to return null instead of BsonNull for we need to return domain value for bindValue; Struct -> BsonDocument transformation should be the only exception; for that reason, I changed the method return type to better align with its purpose.

NathanQingyangXu · 2025-07-11T20:37:14Z

src/main/java/com/mongodb/hibernate/internal/type/MongoArrayJdbcType.java

+            return extractor.getJavaType().wrap(domainObjects, options);
+        } else {
+            return extractor.getJavaType().wrap(array, options);
+        }


The above code changes is to bypass a newly found Hibernate ORM bug. Luckily in this case we can work around.
I created the Hibernate ORM ticket at https://hibernate.atlassian.net/browse/HHH-19596. The bugfix has been merged to both v7 and v6.6.
Just a couple of days we could eliminate the above change after a new v6.6 Hibernate ORM version is released.

I bumped the version to latest v6.6.21 and found the issue has been solved.

vbabanin · 2025-07-21T05:15:40Z

src/integrationTest/java/com/mongodb/hibernate/query/select/Book.java

@@ -43,4 +42,9 @@ class Book {
        this.publishYear = publishYear;
        this.outOfStock = outOfStock;
    }
+
+    @Override
+    public String toString() {


It looks like toString was added mainly to make test output easier to read when troubleshooting. If that’s the goal, it might be worth including all fields so the log clearly reflects the full object state. Otherwise, partial output could still leave us guessing. Should we extend toString to include all fields?

In the current implementation, we use getBooksByIds(int... ids) to go about the retrieval validation. If validation failed, we can easily figure out why based on the id after looking up in the testingBooks list. From my opinion, id alone suffices and we don't need to update the toString() method whenever we modify the entity.

...egrationTest/java/com/mongodb/hibernate/query/select/SortingSelectQueryIntegrationTests.java

src/main/java/com/mongodb/hibernate/jdbc/MongoPreparedStatement.java

…t.java Co-authored-by: Viacheslav Babanin <[email protected]>

… as null"

# Conflicts: # src/integrationTest/java/com/mongodb/hibernate/query/Book.java # src/integrationTest/java/com/mongodb/hibernate/query/select/SortingSelectQueryIntegrationTests.java

…sertionIntegrationTests

NathanQingyangXu · 2025-07-28T20:39:15Z

...ionTest/java/com/mongodb/hibernate/embeddable/StructAggregateEmbeddableIntegrationTests.java

    void testReadNestedValuesMissingFields() {
        var insertResult = mongoCollection.insertOne(
                BsonDocument.parse(
                        """
                        {
                            _id: 1,
-                            nested1: {},
+                            nested1: {a: 0},


Though Hibernate ORM v6.6.23 fixed the bug, we still can't make this testing pass without making the above change. The reason is Single class's a field is of primitive type, and without hibernate.create_empty_composites.enabled enabled (deprecated since v7 and we decided not to use it though we are currently relying on v6.6), we could only setting default value to avoid exception.

So here is the dilemma. If a primitive type field is entity field, null or missing field would end up with default value assigned (due to JDBC ResultSet API spec); but if it shows up in a Struct field, exception would be thrown (Embeddable has its own setter logic different from JDBC API). Given the boolean flag has been removed since v7, it seems we can do nothing here.

NathanQingyangXu · 2025-07-29T18:48:43Z

gradle/libs.versions.toml

@@ -18,7 +18,7 @@ assertj = "3.27.3"
 google-errorprone-core = "2.36.0"
 nullaway = "0.12.4"
 jspecify = "1.0.0"
-hibernate-orm = "6.6.9.Final"
+hibernate-orm = "6.6.23.Final"


this release will fix the annoying https://hibernate.atlassian.net/browse/HHH-19575 bug.

…tionTests

allow null value persistence and retrieval

79ad91e

NathanQingyangXu commented Jul 11, 2025

View reviewed changes

NathanQingyangXu marked this pull request as ready for review July 11, 2025 20:38

NathanQingyangXu requested a review from a team as a code owner July 11, 2025 20:38

NathanQingyangXu requested review from katcharov and removed request for a team July 11, 2025 20:38

Merge branch 'main' into HIBERNATE-48

1c1d888

katcharov requested a review from vbabanin July 11, 2025 22:07

NathanQingyangXu added 3 commits July 15, 2025 07:13

bump hibernte version to v6.6.21 to enjoy some bugfixings

4eb8a5d

Merge remote-tracking branch 'origin/HIBERNATE-48' into HIBERNATE-48

d3cdce8

fix static check issue

680a5c4

NathanQingyangXu requested a review from stIncMale July 16, 2025 13:33

Merge branch 'main' into HIBERNATE-48

5a58a8c

NathanQingyangXu mentioned this pull request Jul 21, 2025

ternary nullness logic implementation #113

Closed

vbabanin reviewed Jul 22, 2025

View reviewed changes

NathanQingyangXu and others added 3 commits July 22, 2025 10:13

Update src/main/java/com/mongodb/hibernate/jdbc/MongoPreparedStatemen…

ebf9094

…t.java Co-authored-by: Viacheslav Babanin <[email protected]>

change per code review comments

092454e

Merge remote-tracking branch 'origin/HIBERNATE-48' into HIBERNATE-48

61bd725

NathanQingyangXu requested a review from vbabanin July 22, 2025 14:16

vbabanin previously approved these changes Jul 22, 2025

View reviewed changes

NathanQingyangXu added 2 commits July 28, 2025 09:32

bump hibernate ORM to 6.6.23 to enjoy bugfix of "empty struct loading…

5da1980

… as null"

Merge branch 'main' into HIBERNATE-48

5634537

# Conflicts: # src/integrationTest/java/com/mongodb/hibernate/query/Book.java # src/integrationTest/java/com/mongodb/hibernate/query/select/SortingSelectQueryIntegrationTests.java

NathanQingyangXu dismissed vbabanin’s stale review via db8489e July 28, 2025 14:53

merge latest main and resolve conflict

e1fff5f

NathanQingyangXu force-pushed the HIBERNATE-48 branch from db8489e to e1fff5f Compare July 28, 2025 15:08

allow literal null value usage in AbstractMqlTranslator and update In…

898909a

…sertionIntegrationTests

NathanQingyangXu commented Jul 28, 2025

View reviewed changes

NathanQingyangXu commented Jul 29, 2025

View reviewed changes

fix a legacy minor comment defect in StructAggregateEmbeddableIntegra…

0b2286b

…tionTests

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

allow null value persistence and retrieval #111

allow null value persistence and retrieval #111

Uh oh!

NathanQingyangXu commented Jul 11, 2025 •

edited

Loading

Uh oh!

NathanQingyangXu Jul 11, 2025 •

edited

Loading

Uh oh!

NathanQingyangXu Jul 11, 2025

Uh oh!

NathanQingyangXu Jul 11, 2025

Uh oh!

NathanQingyangXu Jul 16, 2025

Uh oh!

vbabanin Jul 21, 2025

Uh oh!

NathanQingyangXu Jul 22, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

NathanQingyangXu Jul 28, 2025 •

edited

Loading

Uh oh!

NathanQingyangXu Jul 29, 2025

Uh oh!

Uh oh!


		private static final Set<String> COMPARISON_OPERATOR_NAMES_SUPPORTED =
		Set.of("$eq", "$ne", "$gt", "$gte", "$lt", "$lte");

allow null value persistence and retrieval #111

Are you sure you want to change the base?

allow null value persistence and retrieval #111

Uh oh!

Conversation

NathanQingyangXu commented Jul 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

NathanQingyangXu Jul 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

NathanQingyangXu Jul 11, 2025

Choose a reason for hiding this comment

Uh oh!

NathanQingyangXu Jul 11, 2025

Choose a reason for hiding this comment

Uh oh!

NathanQingyangXu Jul 16, 2025

Choose a reason for hiding this comment

Uh oh!

vbabanin Jul 21, 2025

Choose a reason for hiding this comment

Uh oh!

NathanQingyangXu Jul 22, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

NathanQingyangXu Jul 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

NathanQingyangXu Jul 29, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

NathanQingyangXu commented Jul 11, 2025 •

edited

Loading

NathanQingyangXu Jul 11, 2025 •

edited

Loading

NathanQingyangXu Jul 28, 2025 •

edited

Loading