We're using https://github.com/vladmihalcea/db-util which is a great tool, but we're facing a challenge regarding Session cache, that as mentioned in another question, it can't be disabled .
So, the following test fails because findOne
fetches the object from the session cache:
@Test
public void validateQueries() {
// when (scenario definition)
TestObject testObject = new TestObject(1L);
repository.save(testObject);
SQLStatementCountValidator.reset();
repository.findOne(1L);
SQLStatementCountValidator.assertSelectCount(1);
}
There's a workaround calling entityManager.clear()
every time SQLStatementCountValidator.reset()
is called.
Now, the workaround is fine, but error-prone because now we have to inject EntityManager as a dependency of our tests and remember to call entityManager.clear()
after saving all the objects that represent our scenario.
Questions
Here you can check the log statements (the last one)
09:59.956 [main] [TRACE] o.h.e.i.AbstractSaveEventListener - Transient instance of: TestObject
09:59.957 [main] [TRACE] o.h.e.i.DefaultPersistEventListener - Saving transient instance
09:59.962 [main] [TRACE] o.h.e.i.AbstractSaveEventListener - Saving [TestObject#<null>]
Hibernate:
insert
into
test_object
(id, creation_time, "update_time", "name")
values
(null, ?, ?, ?)
10:00.005 [main] [TRACE] o.h.e.i.AbstractFlushingEventListener - Flushing session
10:00.005 [main] [DEBUG] o.h.e.i.AbstractFlushingEventListener - Processing flush-time cascades
10:00.007 [main] [DEBUG] o.h.e.i.AbstractFlushingEventListener - Dirty checking collections
10:00.007 [main] [TRACE] o.h.e.i.AbstractFlushingEventListener - Flushing entities and processing referenced collections
10:00.011 [main] [TRACE] o.h.e.i.AbstractFlushingEventListener - Processing unreferenced collections
10:00.011 [main] [TRACE] o.h.e.i.AbstractFlushingEventListener - Scheduling collection removes/(re)creates/updates
10:00.011 [main] [DEBUG] o.h.e.i.AbstractFlushingEventListener - Flushed: 0 insertions, 0 updates, 0 deletions to 1 objects
10:00.011 [main] [DEBUG] o.h.e.i.AbstractFlushingEventListener - Flushed: 0 (re)creations, 0 updates, 0 removals to 0 collections
10:00.015 [main] [TRACE] o.h.e.i.AbstractFlushingEventListener - Executing flush
10:00.015 [main] [TRACE] o.h.e.i.AbstractFlushingEventListener - Post flush
10:02.780 [main] [TRACE] o.h.e.i.DefaultLoadEventListener - Loading entity: [TestObject#1]
10:08.439 [main] [TRACE] o.h.e.i.DefaultLoadEventListener - Attempting to resolve: [TestObject#1]
10:08.439 [main] [TRACE] o.h.e.i.DefaultLoadEventListener - Resolved object in session cache: [TestObject#1]
com.vladmihalcea.sql.exception.SQLSelectCountMismatchException: Expected 1 statements but recorded 0 instead!
This is how the workaround code looks like:
@Test
public void validateQueries() {
// when (scenario definition)
TestObject testObject = new TestObject(1L);
repository.save(testObject);
entityManager.clear();
SQLStatementCountValidator.reset();
repository.findOne(1L);
SQLStatementCountValidator.assertSelectCount(1);
}
Each test should manage transactions. So, you should remove the @Transactional
annotation you added at the class level.
So, you inject a TransactionTemplate
bean:
@Autowired
private TransactionTemplate transactionTemplate;
And, then you save the entity in one transaction:
@Test
public void validateQueries() {
try {
transactionTemplate.execute((TransactionCallback<Void>) transactionStatus -> {
TestObject testObject = new TestObject(1L);
repository.save(testObject);
return null;
});
} catch (TransactionException e) {
LOGGER.error("Failure", e);
}
SQLStatementCountValidator.reset();
repository.findOne(1L);
SQLStatementCountValidator.assertSelectCount(1);
}
You can extract the transaction handling logic in a base class method to simplify the exception handling.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.