This commit improves the Javadoc regarding transactional semantics for
@TransactionalEventListener methods invoked in the AFTER_COMMIT,
AFTER_ROLLBACK, or AFTER_COMPLETION phase. Specifically, the
documentation now points out that interactions with the underlying
transactional resource will not be committed in those phases.
Closes gh-26974
Issues gh-25038 and gh-25618 collectively introduced a regression for
thread-scoped and transaction-scoped beans.
For example, given a thread-scoped bean X that depends on another
thread-scoped bean Y, if the names of the beans (when used as map keys)
end up in the same bucket within a ConcurrentHashMap AND an attempt is
made to retrieve bean X from the ApplicationContext prior to retrieving
bean Y, then the use of Map::computeIfAbsent in SimpleThreadScope
results in recursive access to the same internal bucket in the map.
On Java 8, that scenario simply hangs. On Java 9 and higher,
ConcurrentHashMap throws an IllegalStateException pointing out that a
"Recursive update" was attempted.
In light of these findings, we are reverting the changes made to
SimpleThreadScope and SimpleTransactionScope in commits 50a4fdac6e and
148dc95eb1.
Closes gh-25801
PR gh-25038 introduced regressions in SimpleThreadScope and
SimpleTransactionScope in Spring Framework 5.2.7. Specifically, if a
thread-scoped or transaction-scoped bean has a dependency on another
thread-scoped or transaction-scoped bean, respectively, a
ConcurrentModificationException will be thrown on Java 11 or higher.
The reason is that Java 11 introduced a check for concurrent
modification in java.util.HashMap's computeIfAbsent() implementation,
and such a modification can occur when a thread-scoped bean is being
created in order to satisfy a dependency of another thread-scoped bean
that is currently being created.
This commit fixes these regressions by switching from HashMap to
ConcurrentHashMap for the instance maps in SimpleThreadScope and
SimpleTransactionScope.
Closes gh-25618
Prior to this commit, the TransactionAttributeSourceClassFilter
filtered out PlatformTransactionManager but not
ReactiveTransactionManager implementations.
TransactionAttributeSourceClassFilter now filters out any
TransactionManager implementation, covering both imperative and
reactive transaction managers.
This commit introduces tests for the status quo in production support
for multiple transaction managers registered as @Primary and via the
TransactionManagementConfigurer API.
See gh-24869
This commit renames the Runnable variant to executeWithoutResult
and uses a Consumer<TransactionStatus> parameter for better
consistency with TransactionCallbackWithoutResult.
Closes gh-23724
TransactionOperator.as(Mono) now no longer short-cuts via a Flux.next() but provides an implementation via Mono.usingWhen(…).
The short-cut previously issued a cancellation signal to the transactional Mono causing the transaction cleanup to happen without a handle for synchronization.
Using Mono.usingWhen(…) initiates transaction cleanup when the Mono completes eliminating the need for cancellation of the transactional Publisher.
This change does not fully fix gh-23304 but it softens its impact because TransactionalOperator.transactional(Mono) avoids cancellation.
As a follow-up of gh-22915, the purpose of this commit is to improve
Coroutines programmatic transaction API to make it more consistent with
the Java one and more idiomatic.
For suspending functions, this commit changes the
TransactionalOperator.transactional extension with a suspending lambda
parameter to a TransactionalOperator.executeAndAwait one which is
conceptually closer to TransactionalOperator.execute Java API so more
consistent.
For Flow, the TransactionalOperator.transactional extension is correct
but would be more idiomatic as a Flow extension.
This commit also adds code samples to the reference documentation.
Closes gh-23627
