This commit removes the proxyTargetAware attribute from @MockitoSpyBean
while keeping the underlying feature in tact (i.e., transparent
verification for spies created via @MockitoSpyBean).
Closes gh-33775
Prior to this commit, SpringAopBypassingVerificationStartedListener
provided partial support for transparent verification for Mockito spies
created via @MockitoSpyBean when the spy is wrapped in a Spring AOP
proxy. However, attempting to actually verify invocations for a spy
resulted in an exception from Mockito since MockUtil.isMock() returned
false in such scenarios.
This commit addresses that by introducing a SpringMockResolver that
resolves mocks by walking the Spring AOP proxy chain until the target
or a non-static proxy is found.
SpringMockResolver is automatically registered whenever the spring-test
JAR is on the classpath, allowing Mockito to transparently resolve mocks
wrapped in Spring AOP proxies.
Closes gh-33774
This commit removes all references to the Resin Servlet container, as it
is not supported as of Spring Framework 6.0 because we require a
JakartaEE baseline.
Closes gh-33772
A workaround was added for the Resin Servlet container in gh-13937.
This avoids attempting to delete parts that are not named, because the
`part.delete()` call would fail for non-file entries. This can be
problematic for files that are unnamed as they might not be removed by
the Framework.
This commit removes this workaround as Resin is not supported anymore.
Fixes gh-33511
Update `PathMatchingResourcePatternResolver` so that in addition to
searching the `java.class.path` system property for classpath enties,
it also searches the `MANIFEST.MF` files from within those jars.
Prior to this commit, the `addClassPathManifestEntries()` method
expected that the JVM had added `Class-Path` manifest entries to the
`java.class.path` system property, however, this did not always happen.
The updated code now performs a deep search by loading `MANIFEST.MF`
files from jars discovered from the system property. To deal with
potential performance issue, loaded results are also now cached.
The updated code has been tested with Spring Boot 3.3 jars extracted
using `java -Djarmode=tools`.
See gh-33705
Prior to this commit, the SpEL Tokenizer threw an IllegalStateException
when an unsupported character was detected in a SpEL expression;
however, the Javadoc for ExpressionParser.parseExpression() states that
it throws a ParseException if an exception occurred during parsing.
This commit addresses that issue by throwing a SpelParseException for
an unsupported character in a SpEL expression, using a new
UNSUPPORTED_CHARACTER enum constant in SpelMessage.
Closes gh-33767
This commit fixes the issue where the regex pattern in the reference documentation
was not rendering correctly for the `/projects/{project:[a-z]+}/versions` mapping.
Closes gh-33766
Prior to this commit, the instrumentation of the processing of JMS
messages would happen a different levels of the hierarchy, accounting
for alli known implementations, `SimpleMessageListenerContainer` and
`DefaultMessageListenerContainer` as well as various use cases and
`MessageListener` variants.
Unfortunately, this instrumentation could lead to observing JMS
processing twice in some cases, and would not be consistent about the
scope of what's observed.
This commit moves the instrumentation basics into the
`AbstractMessageListenerContainer` but leaves the actual observation
calls to the public implementations.
Fixes gh-33758
This commit updates our Date/Time formatting/printing tests to
demonstrate that the use of fallback patterns can help mitigate
locale-based parsing/formatting issues beginning with JDK 20.
The documentation within the tests is intentionally rather thorough for
two reasons:
1. We need to understand exactly what it is we are testing and why the
tests are written that way.
2. We may re-use parts of the documentation and examples in forthcoming
documentation that we will provide to users.
See gh-33151
Prior to this commit, DefaultListableBeanFactory's
determineHighestPriorityCandidate() method sometimes failed to
determine the highest priority candidate if duplicate priority
candidates were detected whose priority was not the highest priority in
the candidate set. In addition, the bean registration order affected
the outcome of the algorithm: if the highest priority was detected
before other duplicate priorities were detected, the algorithm
succeeded in determining the highest priority candidate.
This commit addresses those shortcomings by ignoring duplicate
@Priority values unless the duplication is for the highest priority
encountered, in which case a NoUniqueBeanDefinitionException is still
thrown to signal that multiple beans were found with the same "highest
priority".
Closes gh-33733
Brian Clozel
Sam Brannen
Juergen Hoeller
rstoyanchev
Nicklas Wiegandt
Phillip Webb
Ilya Serkov
Stéphane Nicoll
Sébastien Deleuze
Sehwan Lim
RollW