This change upgrades the spring-framework build to AspectJ 1.7.1 to
ensure compatibility. We remain backward-compatible to at least AJ
1.6.12 (the previous version), and likely back to 1.6.7.
AJ 1.7 allows for weaving Java 7 bytecode, particularly with regard
to invokedynamic instructions, and furthermore makes improvements to
the aspectj language itself to allow users to take advantage of Java 7-
style language features within aspects. See [1] for details.
[1]: http://eclipse.org/aspectj/doc/released/README-170.html
Issue: SPR-10079
This change introduces a new AllEncompassingFormHttpMessageConverter
class that adds JSON and XML converters for individual mime parts of
a multi-part request. The new converter is used in place of the
previously used XmlAwareFormHttpMessageConverter.
Issue: SPR-10055
This commit upgrades the compile dependency on JUnit in the spring-test
module as well as the global test dependency on JUnit for the entire
test suite to 4.11 GA.
In addition, the spring-test module now depends on junit:junit instead
of junit:junit-dep since, as of JUnit 4.11, the "junit" Maven artifact
no longer inlines hamcrest-core.
Issue: SPR-10044
Tiles 3 has modified packages and classes as well as additional
dependencies, notably "tiles-request-api", which is a request/response
abstraction independent of Servlet and JSP APIs.
In order to have both Tiles 2 and Tiles 3 integrations, the source for
the Tiles 3 integration is in a separate project spring-webmvc-tiles3.
The build process merges the compiled Tiles 3 integration classes into
the spring-webmvc module so in effect it contains both the Tiles 2 and
the Tiles 3 integrations.
This change originated as a pull request at spring-framework-issues:
https://github.com/SpringSource/spring-framework-issues/pull/30
And was additionally updated:
1f64be4aa5
Issue: SPR-8825
This is more specific exception raised instead of RestClientException
when the raw HTTP status code received from the server is not one of
the HttpStatus enum values.
Issue: SPR-9406
The MVC namespace and the MVC Java config now allow configuring
CallableProcessingInterceptor and DeferredResultProcessingInterceptor
instances.
Issue: SPR-9914
Before this change View implementations set the response content type
to the fixed value they were configured with.
This change makes it possible to configure a View implementation with
a more general media type, e.g. "application/*+xml", and then set the
response type to the more specific requested media type, e.g.
"application/vnd.example-v1+xml".
Issue: SPR-9807.
Prior to this change, the repackaged spring-asm and spring-cglib jars
were being included wholesale in the spring-core jar, whereas the
intention was to include the unzipped classfiles.
This change ensures that spring-asm and spring-cglib jars are unzipped
on the fly when creating the spring-core jar.
Issue: SPR-9669
CGLIB 3 has been released in order to depend on ASM 4, which Spring now
depends on internally (see previous commit).
This commit eliminates spring-beans' optional dependency on cglib-nodep
v2.2 and instead repackages net.sf.cglib => org.springframework.cglib
much in the same way we have historically done with ASM.
This change is beneficial to users in several ways:
- Eliminates the need to manually add CGLIB to the application
classpath; especially important for the growing number of
@Configuration class users. Java-based configuration functionality,
along with proxy-target-class and method injection features now
work 'out of the box' in Spring 3.2.
- Eliminates the possibility of conflicts with other libraries that
may dependend on differing versions of CGLIB, e.g. Hibernate
3.3.1.ga and its dependency on CGLIB 2.1.3 would easily cause a
conflict if the application were depending on CGLIB 3 for
Spring-related purposes.
- Picks up CGLIB 3's changes to support ASM 4, meaning that CGLIB is
that much less likely to work well in a Java 7 environment due to
ASM 4's support for transforming classes with invokedynamic
bytecode instructions.
On CGLIB and ASM:
CGLIB's own dependency on ASM is also transformed along the way to
depend on Spring's repackaged org.springframework.asm, primarily to
eliminate unnecessary duplication of ASM classfiles in spring-core and
in the process save around 100K in the final spring-core JAR file size.
It is coincidental that spring-core and CGLIB currently depend on the
exact same version of ASM (4.0), but it is also unlikely to change any
time soon. If this change does occur and versions of ASM drift, then
the size optimization mentioned above will have to be abandoned. This
would have no compatibility impact, however, so this is a reasonable
solution now and for the forseeable future.
On a mysterious NoClassDefFoundError:
During the upgrade to CGLIB 3.0, Spring test cases began failing due to
NoClassDefFoundErrors being thrown from CGLIB's DebuggingClassWriter
regarding its use of asm-util's TraceClassVisitor type. previous
versions of cglib-nodep, particularly 2.2, did not cause this behavior,
even though cglib-nodep has never actually repackaged and bundled
asm-util classes. The reason for these NoClassDefFoundErrors occurring
now is still not fully understood, but appears to be due to subtle JVM
bytecode preverification rules. The hypothesis is that due to minor
changes in DebuggingClassWriter such as additional casts, access to
instance variables declared in the superclass, and indeed a change in
the superclass hierarchy, preverification may be kicking in on the
toByteArray method body, at which point the reference to the missing
TraceClassVisitor type is noticed and the NCDFE is thrown. For this
reason, a dummy implementation of TraceClassVisitor has been added to
spring-core in the org.springframework.asm.util package. This class
simply ensures that Spring's own tests never result in the NCDFE
described above, and more importantly that Spring's users never
encounter the same.
Other changes include:
- rename package-private Cglib2AopProxy => CglibAopProxy
- eliminate all 'cglibAvailable' checks, warnings and errors
- eliminate all 'CGLIB2' language in favor of 'CGLIB'
- eliminate all mention in reference and java docs of needing to add
cglib(-nodep) to one's application classpath
Issue: SPR-9669
ASM 4.0 is generally compatibile with Java 7 classfiles, particularly
including 'invokedynamic' instructions. This is important when
considering that Spring's component-scanning support is internally
ASM-based and it is increasingly likely that component classes having
invokedynamic instructions may be encountered and read by ASM.
This upgrade, then, is primarily preventive in nature.
Changes include:
- upgrade from ASM 2.2.3 to ASM 4.0
- adapt to ASM API changes as necessary throughout spring-core,
resulting in no impact to the public Spring API.
- remove dedicated spring-asm module
- use new :spring-core:asmRepackJar task to repackage
org.objectweb.asm => org.springframework.asm as per usual and write
repackaged classes directly into spring-core jar
The choice to eliminate the spring-asm module altogether and instead
inline the repackaged classes directly into spring-core is first to
eliminate an otherwise unnecessary second jar. spring-core has a
non-optional dependency on spring-asm meaning it is always on the
application classpath. This change simplifies that situation by
consoliding two jars into one. The second reason for this choice is in
anticipation of upgrading CGLIB to version 3 and inlining it into
spring-core as well. See subsequent commit for details.
Issue: SPR-9669
Attempting to register a custom MethodFilter with a
StandardEvaluationContext after invoking setMethodResolvers() with a
custom list of MethodResolver instances results in a
NullPointerException. Based on the current documentation in
StandardEvaluationContext it is unclear what the expected behavior
should be, but either the implementation is broken, or the use case is
unsupported. In either case, allowing a NullPointerException to be
thrown is inappropriate.
This commit documents the fact that the SpEL MethodFilter is intended to
be used with the ReflectiveMethodResolver. Furthermore,
StandardEvaluationContext.registerMethodFilter() now throws an
IllegalStateException if the user attempts to set a filter after having
registered a custom set of resolvers.
Issue: SPR-9621
The Spring Expression Language currently supports nested single quotes
within expressions but not nested double quotes.
The SpEL tokenizer has been modified to support nested double quotes in
the same way it supports single quotes. A sequence of two double quotes
will now be replaced by one when evaluated.
Extra error handling has also been added to report when invalid escaping
is encountered, since SpEL does not support escaping with backslash.
Issue: SPR-9620
SpEL typically supports logical operators for boolean expressions
consistent with standard Java language syntax. However, the operators
for logical AND and logical OR are currently only supported as textual
operators. In other words, SpEL does not support the use of && and || as
logical operators.
The SpEL tokenizer has now been modified to recognize && and || as
symbolic boolean operators. The parser has been modified to allow the
use of either the textual or symbolic operators.
Issue: SPR-9614
- renamed resolveParameterizedReturnType() to
resolveReturnTypeForGenericMethod()
- fleshed out Javadoc for resolveReturnType() and
resolveReturnTypeForGenericMethod() regarding declaration of formal
type variables
- improved wording in log statements and naming of local variables
within resolveReturnTypeForGenericMethod()
Issue: SPR-9493
Prior to this commit null literals in SpEL expressions had to be
specified as "null" (i.e., all lowercase).
With this commit null literals in SpEL expressions are interpreted in a
case-insensitive manner, analogous to the current support for boolean
literals.
Issue: SPR-9613
When attempting to parse an Integer literal expression such as
42.toString(), SpEL currently throws a SpelParseException with a message
similar to: "EL1041E:(pos 3): After parsing a valid expression, there is
still more data in the expression: 'toString'". The problem here is that
'3.' is currently considered a valid number (including the dot).
However, SpEL succeeds at parsing an equivalent expression for a Double
literal such as 3.14.isInfinite().
To address this issue, the SpEL Tokenizer no longer consumes the
trailing '.' on an integer as part of the integer. So '3.foo()' will now
be parsed as '3' '.' 'foo()' and not '3.' 'foo()' -- which was what
prevented parsing of method invocations on integers. To keep the change
simple, the parser will no longer handle real numbers of the form
'3.e4'. From now on they must include the extra 0 (i.e., '3.0e4').
Issue: SPR-9612
