The modifications to DefaultMockMvcBuilder performed in conjunction
with SPR-12553 introduced a breaking change: the WebApplicationContext
supplied to DefaultMockMvcBuilder's constructor was *always* stored in
the ServletContext as the root WebApplicationContext, overwriting a
root WebApplicationContext that had been set by the user or by the
Spring TestContext Framework (TCF) -- for example, in
AbstractGenericWebContextLoader. Consequently, the changes in SPR-12553
cause tests that use @ContextHierarchy to fail if web components rely
on the correct WebApplicationContext being stored under the
WebApplicationContext#ROOT_WEB_APPLICATION_CONTEXT_ATTRIBUTE key.
This commit reverts the breaking changes introduced in SPR-12553: if
the root WebApplicationContext has already been set in the
ServletContext of the WebApplicationContext supplied to
DefaultMockMvcBuilder, no action is taken.
Furthermore, this commit introduces new code to address the initial
intent of SPR-12553. Specifically, if the root WebApplicationContext
has NOT been set in the ServletContext of the WebApplicationContext
supplied to DefaultMockMvcBuilder, the application context hierarchy
will be traversed in search of the root WebApplicationContext, and the
root WebApplicationContext will then be stored under the
ROOT_WEB_APPLICATION_CONTEXT_ATTRIBUTE key.
Issue: SPR-13075, SPR-12553
(cherry picked from commit f6d2fe471a)
Since Spring 4.1, a CacheResolver may be configured to customize the way
the cache(s) to use for a given cache operation are retrieved. Since a
CacheResolver implementation may not use the cache names information at
all, this attribute has been made optional.
However, a fix was still applied, preventing a Cache operation without a
cache name to be defined properly. We now allow this valid use case.
Issue: SPR-13081
Previously, a custom `DestinationResolver` was not made available
to the underlying `MessageListener`: if a reply needs to be sent, the
default `DestinationResolver` was used.
This commit ensures that if a custom `DestinationResolver` has been set,
it is also configured accordingly for replies.
Issue: SPR-12927
Prior to this commit, the `ResourceHttpRequestHandler` would not
properly handle HTTP requests to **directories contained in JARs**.
This would result in HTTP 500 errors, caused by `FileNotFoundException`
or `NullPointerException`.
This can be tracked to webapp ClassLoader implementations in servlet
containers:
* in Jetty9x, fetching a directory within a JAR as a `Resource` and
getting its InputStream work fine, but attempting to `close()` it
results in a NullPointerException as the underlying stream is null.
* In Tomcat6x, one cannot fetch an InputStream for the same `Resource`
as it throws a FileNotFoundException.
This change adds more try/catch clauses and catches more Exception so as
to result in HTTP 200 OK responses instead of server errors. While this
is inconsistent because the same code path would result in HTTP 404 with
existing directories on the file system, there's no other simple way to
make those checks for resources contained in JARs.
Issue: SPR-12999
Prior to this commit, the ResourceUrlEncodingFilter would fail with a
StringIndexOutOfBoundsException when:
* the current request has a servlet context
* the URL to encode is relative and is shorter than the context value
This change defensively checks for those lengths and delegates to the
parent implementation if necessary.
Issue: SPR-13018
When used in combination with GzipResourceResolver, the
CachingResourceResolver does not properly cache results, since it only
takes the request path as a input for cache key generation.
Here's an example of that behavior:
1. an HTTP client requests a resource with `Accept-Encoding: gzip`, so
the GzipResourceResolver can resolve a gzipped resource.
2. the configured CachingResourceResolver caches that resource.
3. another HTTP client requests the same resource, but it does not
support gzip encoding; the previously cached gzipped resource is still
returned.
This commit uses the presence/absence of gzip encoding support as an
input in cache keys generation.
Issue: SPR-12982
Also refines logIntrospectionFailure to just log at debug level for meta-annotation introspection failures.
Issue: SPR-12889
(cherry picked from commit e78b086)
Prior to this commit, AntPathMatcher would not correctly combine a path
that ends with a separator with a path that starts with a separator.
For example, `/foo/` + `/bar` combined into `/foo//bar`.
Specifically, this commit:
- Removes the duplicated separator in combined paths
- Improves RequestMappingInfo's toString() representation
- Fixes Javadoc formatting in AntPathMatcher
- Polishes AntPathMatcherTests
- Polishes Javadoc in AbstractRequestCondition
Issue: SPR-12975
(cherry picked from commit a80d0e87a4)
