Prior to this change, the ShallowEtagHeaderFilter would use a
ResizableByteArrayOutputStream to internally write data and calculate
the ETag. While that implementation is faster than the regular
ByteArrayOutputStream (since it has a better strategy for growing the
internal buffer), a lot of buffer copying/writing still happens.
This change adds a new FastByteArrayOutputStream implementation that
internally uses a LinkedList<Byte[]> to store the content. So when
writing bytes to that OutputStream implementation, new byte[] are
added to the list when the previous ones are full. This saves most
of the instantiating/copying operations.
Note that new methods were added in DigestUtils to allow usage of
Streams instead of byte[], which is more efficient in our case.
Fixes#653
Issue: SPR-12081
This commit introduces test methods in CollectionFactoryTests that
demonstrate how the APIs for createCollection() and createMap() are not
type-safe, specifically regarding the use of generics, raw types, and
casting.
SPR-12483 introduced automatic type conversion support for EnumSet and
EnumMap. However, the corresponding changes in CollectionFactory
contradict the existing contract for the "create approximate" methods
by creating a copy of the supplied set or map, thereby potentially
including elements in the returned collection when the returned
collection should in fact be empty.
This commit addresses this issue by ensuring that the collections
returned by createApproximateCollection() and createApproximateMap()
are always empty.
Furthermore, this commit improves the Javadoc throughout the
CollectionFactory class.
Issue: SPR-12533
This commit introduces test methods in CollectionFactoryTests that
demonstrate how the APIs for createApproximateCollection() and
createApproximateMap() are not type-safe, specifically regarding the use
of generics, raw types, and casting.
JUnit 4.12 created a 'public' version of the previously 'internal'
AssumptionViolatedException.
This commit upgrades Spring's Assume class to use this new public
version.
This commit introduces superfluous casts in CollectionFactory to
address a bug in Eclipse 4.4.1 that prevents the code from compiling
within Eclipse. Specifically, without these casts Eclipse displays the
following error for use of the EnumSet.copyOf() and EnumSet.noneOf()
methods:
Type mismatch: cannot convert from EnumSet<Enum<Enum<E>>> to
Collection<E>
Prior to this commit, JOptCommandLinePropertySource prevented the
possibility of non-String option arguments. This effectively prevents
the use of JOpt's #ofType support (which allows specifying custom
argument types).
Now, non-String arguments are detected and converted to strings as
necessary. JOpt's #ofType now works as expected. A test has been added
to cover this case.
Prior to this commit, the ResolvableType static cache was holding a lot
of duplicates for simple types. We are using too much metadata to compute
the key when the class has no generic information. so setFoo(String foo)
and setBar(String bar) would result in two entries in the cache because
the TypeProvider is different. On a very simple application 65% of the
entries in the cache were duplicate.
When the type is a Class with no generic information, the ResolvableType
instance is a simple wrapper around it so we might just as well not cache
it at all as the cost of finding it back from the cache is higher than
creating that simple wrapper.
This commit adds an explicit check; if the type is a simple Class we just
return a "resolved" ResolvableType instance for it. On a few test cases,
this reduces the size of the cache by 85%
Issue: SPR-12275
Prior to this commit, `AntPathMatcher.extractPathWithinPattern` would
not process correctly `**` patterns and would only match *one* path
segment in the given path.
This commit changes `extractPathWithinPattern` to allow multiple path
segments to be matched against a single `**` pattern segment.
Issue: SPR-10515
Update PathMatchingResourcePatternResolver to include additional
protected methods that can be used by subclasses to optimize which
JARs are searched.
Issue: SPR-12231
Update SystemEnvironmentPropertySource to attempt optimized Map lookups
first, and only fall-back to the defensive SecurityManager safe-mode
if these fail.
Issue: SPR-12224
Per the Javadoc for the SocketUtils() constructor, SocketUtils can be
instantiated as a Spring Bean in XML configuration files; however,
SocketUtils is currently abstract which prevents such usage.
This commit removes the 'abstract' declaration thereby allowing
SocketUtils to be instantiated as a Spring bean.
Issue: SPR-12169
Prior to this commit, when there was a lot of entries in the
ResolvableType.cache HashMap, getting a simple value could
take a lot of time due to a lot of calls to ResolvableType.equals().
ResolvableType.equals() used this.type, getSource(),
this.variableResolver.getSource() and this.componentType, but
ResolvableType.hashCode() used only this.type.
With this commit, ResolvableType.hashCode() now uses the same
fields than ResolvableType.equals().
Performance on the spring-resolvabletype-benchmark project:
- 8000 us before this commit
- 120 us with this commit
Issue: SPR-12122
Rework the @PropertySource parsing logic recently changed in commit
7c608886 to deal with the same source appearing on a @Configuration
class and an @Import class.
Processing now occurs in a single sweep, with any previously added
sources being converted to a CompositePropertySource.
Issue: SPR-12115
Prior to this commit, there was no declarative mechanism for a custom
TestExecutionListener to be registered as a default
TestExecutionListener.
This commit introduces support for discovering default
TestExecutionListener implementations via the SpringFactoriesLoader
mechanism. Specifically, the spring-test module declares all core
default TestExecutionListeners under the
org.springframework.test.context.TestExecutionListener key in its
META-INF/spring.factories properties file, and third-party frameworks
and developers can contribute to the list of default
TestExecutionListeners in the same manner.
- AbstractTestContextBootstrapper uses the SpringFactoriesLoader to
look up the class names of all registered default
TestExecutionListeners and sorts the instantiated listeners using
AnnotationAwareOrderComparator.
- DefaultTestContextBootstrapper and WebTestContextBootstrapper now
rely on the SpringFactoriesLoader mechanism for finding default
TestExecutionListeners instead of hard coding fully qualified class
names.
- To ensure that default TestExecutionListeners are registered in the
correct order, each can implement Ordered or declare @Order.
- AbstractTestExecutionListener and all default TestExecutionListeners
provided by Spring now implement Ordered with appropriate values.
- Introduced "copy constructors" in MergedContextConfiguration and
WebMergedContextConfiguration
- SpringFactoriesLoader now uses AnnotationAwareOrderComparator
instead of OrderComparator.
Issue: SPR-11466
Replace references to the old RFC 2616 (HTTP 1.1) with references
to the new RFCs 7230 to 7235.
This commit also deprecates:
- HttpStatus.USE_PROXY
- HttpStatus.REQUEST_ENTITY_TOO_LARGE in favor of HttpStatus.PAYLOAD_TOO_LARGE
- HttpStatus.REQUEST_URI_TOO_LONG in favor of HttpStatus.URI_TOO_LONG
Issue: SPR-12067
Prior to this commit, given an enum which implements some interface,
GenericConversionService would select the String -> Enum converter even
if a converter for String -> SomeInterface was registered. This also
affected converters that were registered for String ->
SomeBaseInterface, when SomeInterface extended SomeBaseInterface.
This change modifies the behavior of the private method
getClassHierarchy() by placing Enum.class as late as possible, pretty
much the same way as Object.class is handled.
Issue: SPR-12050
Make it possible to use a ListenableFuture with Java 8
lambda expressions, using a syntax like
listenableFuture.addCallback(() -> ..., () -> ...);
Issue: SPR-11820