Spring Data JPA difference between findBy / findAllBy

To illustrate the difference lets look at the two functions:

1. Set<Policy> findAllByRoleIn(Iterable<Role> role);

2. Set<Policy> findByRoleIn(Iterable<Role> role);

The query generated by 1st function:

1.  select policy.id, policy.role from policy where (policy.role in (? , ? , ? , ?))

The query generated by 2nd function:

2. select policy.id, policy.role from policy where (policy.role in (? , ? , ? , ?))

Conclusion: Clearly, if we look at the queries generated by both functions. We can clearly see, there is no difference between the two function definitions, they execute exactly the same query.

No, there is no difference between them, they will execute exactly the same query, the All part is ignored by Spring Data when deriving the query from the method name. The only important bit is the By keyword, anything following it is treated as a field name (with the exception of other keywords like OrderBy which incidentially can lead to some strange looking method names like findAllByOrderByIdAsc).

This means something like this is perfectly valid:

List<SomeEntity> findAnythingYouWantToPutHereBySomeCondition();

And will execute exactly the same SQL query as:

List<SomeEntity> findBySomeCondition();

or

List<SomeEntity> findAllBySomeCondition();

The documentation for the 2.3.6 release of Spring Data discusses this feature:

Any text between find (or other introducing keywords) and By is considered to be descriptive unless using one of the result-limiting keywords such as a Distinct to set a distinct flag on the query to be created or Top/First to limit query results.

The purpose of feature was explained in a blog post about the then-upcoming 2.0 release of Spring Data:

Spring Data’s method parsing uses prefix keywords like find, exists, count, and delete and a terminating By keyword. Everything you put in between find and By makes your method name more expressive and does not affect query derivation.