Ensure that property source locations are processed in the same order
regardless if the 'name' attribute is set or not.
Prior to this commit multiple locations from a `@PropertySource` with
a name were added to a `CompositePropertySource` in such a way that
the first location would take precedence. This has now been reversed
for consistence with unnamed `@PropertySource`s
Issue: SPR-10820
Update `AbstractApplicationEventMulticaster` to support `EventObjects`
with a null source (which can happen if they have been serialized).
Issue: SPR-10945
Update the `CacheKey` class used by `ReflectivePropertyAccessor` to
include if the target object is class. The prevents an incorrect cache
hit from being returned when a property with the same name is read on
both an object and its class. For example:
#{class.name}
#{name}
Issue: SPR-10486
Prior to this commit, the uniqueness check for @ContextConfiguration
attributes within a @ContextHierarchy was performed at a single test
class level instead of against the merged configuration for all test
class levels in the test class hierarchy.
This commit addresses this issue by moving the uniqueness check
algorithm from resolveContextHierarchyAttributes() to
buildContextHierarchyMap() within ContextLoaderUtils.
Issue: SPR-10997
Update getValue(EvaluationContext context, Object rootObject,
Class<T> desiredResultType) to propagate the EvaluationContext to
ExpressionUtils.
Issue: SPR-10953
Update TypeConverterDelegate to bypass conversion when the target type
is Object.class. Prior to this commit converting a single element array
to an Object would result in the element being returned, rather than
the array itself.
Issue: SPR-10996
Relax the method search algorithm used by `ReflectivePropertyAccessor`
to include methods of the form `getXY()` for properties of the form
`xy`.
Although the JavaBean specification indicates that a property `xy`
should use the accessors `getxY()` and `setxY()`, in practice many
developers choose to have an uppercase first character. The
`ReflectivePropertyAccessor` will now consider these style methods if
the traditional conventions fail to find a match.
Issue: SPR-10716
Restore pringing the payload first and headers second as it has been in
SI but also handle specifically the case where the body is a byte array
to minimize unnecessary "noise" that causes otherwise for STOMP msgs.
The STOMP UNSUBSCRIBE message does not have a destination header so
there is no obvious simple way to do a mapping and the need for it
is not clear yet.
Prior to this commit, @SubscribeEvent@UnsubscribeEvent and
@MessageMapping annotated message handling methods
could only match a strict message destination.
This commit adds a @PathVariable annotation and
updates the message matching/handling process, since
message handling methods can now match PathMatcher-like
destinations and get path variables injected in parameters.
Issue: SPR-10949
This version fixed bugs and improved performance.
New features: "select multiple attributes" and
"Array operations and slicing improved" (a new example
has been added in tests to reflect that).
See https://github.com/jayway/JsonPath/blob/master/README
Issue: SPR-10990
Prior to this commit, @ControllerAdvice annotated beans would
assist all known Controllers, by applying @ExceptionHandler,
@InitBinder, and @ModelAttribute.
This commit updates the @ControllerAdvice annotation,
which accepts now base package names, assignableTypes,
annotations and basePackageClasses.
If attributes are set, only Controllers that match those
selectors will be assisted by the annotated class.
This commit does not change the default behavior when
no value is set, i.e. @ControllerAdvice().
Issue: SPR-10222
This change adds abstractions for opening and managing TCP connections
primarily for use with the STOMP broker support. As one immediate
benefit the change makes the StompBrokerRelayMessageHandler more
easy to test.
This change consolidates TCP-related logic in the StompRelaySession. As
a result the sub-class SystemStompRelaySession now contains only logic
intrinsic to the shared system session.
This change upgrades reactor to 1.0 RC1 and also removes the
reactor-based message channel in favor of the one available from
org.projectreactor:reactor-spring.
Refactor TypeDescriptor to use ResolvableType in order to retain full
generic type information, in the process fixing MultiValueMap support
in MapToMapConverter.
Issue: SPR-9499
Unfortunately, the JCache API changed quite a bit since 0.6. We're building against a snapshot of JCache 0.11 now, tracking its way to final after the Public Review Ballot.
- Support for serialization
- Allow programmatic creation of an array from a given component type
- Allow programmatic creation with given generics
- Extract generics from Class types using Class.getTypeParameters()
- Move TypeVariableResolver to an inner class (and make method private)
- Refine 'resolve()' algorithm
Issue: SPR-10973
In order to help developers avoid surprises from unintended use of
"@Bean Lite Mode" when referencing classes via @ContextConfiguration in
the TestContext framework, this commit duplicates the previous
discussion of "annotated classes" from the Javadoc for
@ContextConfiguration as a tip in the "Context configuration with
annotated classes" section of the Testing chapter in the reference
manual.
Issue: SPR-10971
- Now avoiding NullPointerExceptions in GenericTypeResolver's
resolveReturnTypeForGenericMethod() in case the supplied ClassLoader
is null.
- AutowireUtils.resolveReturnTypeForFactoryMethod() now properly
asserts that the supplied ClassLoader is not null.
- Fixed copy-n-paste errors in Javadoc for
AutowireUtils.resolveReturnTypeForFactoryMethod().
Issue: SPR-10411
This change adds an alternative UUID generation strategy to use by
default in MessageHeaders. Instead of using SecureRandom for each
new UUID, SecureRandom is used only for the initial seed to be
provided java.util.Random. Thereafter the same Random instance is
used instead. This provides improved performance while id's are
still random but less securely so.
Before this change, async result handling on controller methods failed
to observe type-level annotations annotations. The issue was never
noticed until we started supporting type-level @ResponseBody and the
@RestController meta annotation.
Issue: SPR-10905
Includes revisions of MethodParameter and DependencyDescriptor (in particular towards a reference to the containing class). Also refines several ResolvableType method signatures.
Issue: SPR-9965
We're consistently resolving class names now, and the entire algorithm moved from GenericTypeResolver to the internal AutowireUtils helper in the bean factory package.
Issue: SPR-10411