* SPR-9284:
Allow MapToMap conversion to work when the target map does not have a default constructor (as long as a new map copy is not required).
Polish trailing whitespace
Update the ReflectiveMethodResolver and ReflectivePropertyAccessor
to allow methods and properties of java.lang.Class to be resolved
when the target object is a class.
Issue: SPR-9017
Refactor ConfigurationClassEnhancer to allow cglib caching of
generated classes. Prior to this commit each enhanced @Configuration
class would consume permgen space when created.
The CallbackFilter and Callback Types are now defined as static final
members so that they can be shared by all enhancers. Only the
callbackInstances remain specific to a @Configuration class and
these are not used by cglib as part of the cache key.
Issue: SPR-9851
This change ensures that when the Accept and the Producible media types
are equally specific, we use the one from the Accept header, which may
for example carry a different charset.
Remove use of UriTemplate in MockMvcRequestBuilders.
Rely on UriComponentsBuilder instead.
Decode query params before setting them on MockHttpServletRequest.
Add more options to model result matching for the count of errors.
Ignore white spaces and comments when comparing XML.
Aside from minor polishing, this change sets the "systemProperties" and "systemEnvironment" beans at each factory level as well.
Issue: SPR-9756
Issue: SPR-9764
Add "MockRest-" prefix to RequestMatchers and ResponseCreators to
make it easy to find classes with static imports with
Ctrl+Shift+T "MockRest" - similar to the "MockMvc"
prefix on the server-side.
This change fixes further cases under JDK 6 in which setting a bridged
(e.g. String-returning) read method can conflict with an existing
corresponding bridge write method that accepts an Object parameter.
This appears to be a implementation difference between JDKs 6 and 7,
where the JDK 6 Introspector adds bridge methods and JDK 7 does not.
The solution here is to consistently null-out any existing write method
before setting the read method. We were doing this elsewhere in
ExtendedBeanInfo already, but these two changes make the approach
consistent throuhout.
Issue: SPR-8806
Prior to this commit, the Spring TestContext Framework only supported
loading an ApplicationContext in integration tests from either XML or
Java Properties files (since Spring 2.5), and Spring 3.1 introduced
support for loading an ApplicationContext in integration tests from
annotated classes (e.g., @Configuration classes). All of the
ContextLoader implementations used to provide this support load a
GenericApplicationContext. However, a GenericApplicationContext is not
suitable for testing a web application since a web application relies on
an implementation of WebApplicationContext (WAC).
This commit makes it possible to integration test Spring-powered web
applications by adding the following functionality to the Spring
TestContext Framework.
- Introduced AbstractGenericWebContextLoader and two concrete
subclasses:
- XmlWebContextLoader
- AnnotationConfigWebContextLoader
- Pulled up prepareContext(context, mergedConfig) from
AbstractGenericContextLoader into AbstractContextLoader to allow it
to be shared across web and non-web context loaders.
- Introduced AnnotationConfigContextLoaderUtils and refactored
AnnotationConfigContextLoader accordingly. These utils are also used
by AnnotationConfigWebContextLoader.
- Introduced a new @WebAppConfiguration annotation to denote that the
ApplicationContext loaded for a test should be a WAC and to configure
the base resource path for the root directory of a web application.
- Introduced WebMergedContextConfiguration which extends
MergedContextConfiguration with support for a baseResourcePath for
the root directory of a web application.
- ContextLoaderUtils.buildMergedContextConfiguration() now builds a
WebMergedContextConfiguration instead of a standard
MergedContextConfiguration if @WebAppConfiguration is present on the
test class.
- Introduced a configureWebResources() method in
AbstractGenericWebContextLoader that is responsible for creating a
MockServletContext with a proper ResourceLoader for the
resourceBasePath configured in the WebMergedContextConfiguration. The
resulting mock ServletContext is set in the WAC, and the WAC is
stored as the Root WAC in the ServletContext.
- Introduced a WebTestExecutionListener that sets up default thread
local state via RequestContextHolder before each test method by using
the MockServletContext already present in the WAC and by creating a
MockHttpServletRequest, MockHttpServletResponse, and
ServletWebRequest that is set in the RequestContextHolder. WTEL also
ensures that the MockHttpServletResponse and ServletWebRequest can be
injected into the test instance (e.g., via @Autowired) and cleans up
thread locals after each test method.
- WebTestExecutionListener is configured as a default
TestExecutionListener before DependencyInjectionTestExecutionListener
- Extracted AbstractDelegatingSmartContextLoader from
DelegatingSmartContextLoader and introduced a new
WebDelegatingSmartContextLoader.
- ContextLoaderUtils now selects the default delegating ContextLoader
class name based on the presence of @WebAppConfiguration on the test
class.
- Tests in the spring-test-mvc module no longer use a custom
ContextLoader to load a WebApplicationContext. Instead, they now
rely on new core functionality provided in this commit.
Issue: SPR-5243
Caching of resovled exceptions introduced in SPR-7703 also introduced a
side effect whereby if exactly one exception was previously cached, any
other exception would appear as a match to the previously matched
@ExceptionHandler method.
This change ensures use of a fresh map when determining matching
@ExceptionHandler methods while also updating the cache.
Issue: SPR-9209
An additional update (after the last commit) of the "includes" and
"isCompatibleWith" methods of MediaType to accomodate wildcards
in media types with a suffix.
Issue: SPR-9841
The "includes" and "isCompatibleWith" methods of MediaType take into
account media types with suffices (e.g. application/soap+xml) including
wildcards with suffices (e.g. application/*+xml). However before this
change, the isWildcardSubtype() method returned true only for subtype
"*". Now a media type such as application/*+xml is also recognized as
having a wildcard subtype.
Issue: SPR-9841
Previously the project that was getting merged from would try to
execute the artifactoryPublish on the build box. This caused issues
since the mavenDescriptor did not exist.
The merge-dist.gradle now disables the artifactoryPublish task if it
exists (it will only exist on the build box).
Issue: SPR-9859, SPR-7951
This commit adds the spring-test-mvc project [1] to the Spring
Framework as part of the spring-test module.
The sources are added as a root-level project called "spring-test-mvc"
instead of under "spring-test" because the new sources need to be
compiled with Servlet 3 while the current "spring-test" sources require
Servlet 2.5 and the Eclipse IDE does not support having different
classpaths for the same project.
The Gradle build produces a single spring-test jar that contains
sources from both "spring-test" and "spring-test-mvc". This merge is
made possible through merge-dist.gradle as follows:
- jar tasks of the "from" project execute tasks of the "to" project
- "to" project is added to the classpath of the "from" project
- "to" project pom is updated with entries from the "from" project
For further details see documentation in merge-dist.gradle.
Special thanks to everyone who contributed to the initial development
of the Spring MVC Test framework:
Arjen Poutsma <poutsma@mac.com>
Craig Walls <cwalls@vmware.com>
Frans Flippo <fransflippo@utopia.orange11.nl>
Harry Lascelles <harry@firstbanco.com>
Irfan <mail.urfi@gmail.com>
Jörg Rathlev <joerg.rathlev@s24.com>
Keesun Baik <whiteship2000@gmail.com>
Keesun Baik <whiteship@epril.com>
Matthew Reid <matthew.reid@nakedwines.com>
Nils-Helge Garli Hegvik <Nils-Helge.Hegvik@telenor.com>
Rob Winch <rwinch@vmware.com>
Scott Frederick <sfrederick@vmware.com>
Sven Filatov <sven.filatov@gmail.com>
Thomas Bruyelle <thomas.bruyelle@gmail.com>
youngm <youngm@gmail.com>
[1]: https://github.com/SpringSource/spring-test-mvc
Issue: SPR-9859, SPR-7951
This should be significantly faster than our standard algorithm, for a very common case. Motivated by SPR-9802, even if the fix there uses a different approach, with transaction name determination not calling getMostSpecificMethod at all anymore.
Issue: SPR-9802