[[readme-1_1]] = AspectJ 1.1 _© Copyright 2002 Palo Alto Research Center, Incorporated, 2003 Contributors. All rights reserved._ This is the initial release of AspectJ 1.1. It includes a small number of new language features as well as major improvements to the functionality of the tools. This document describes the differences between AspectJ versions 1.1 and 1.0.6. Users new to AspectJ need only read the link:progguide/index.html[AspectJ Programming Guide] since it describes the 1.1 language. Users familiar with AspectJ 1.0 may find this document a quicker way to learn what changed in the language and tools, and should use it as a guide for porting programs from 1.0 to 1.1. This document first summarizes changes from the 1.0 release in * xref:#language[the language], * xref:#compiler[the compiler], * xref:#tools[the support tools], * xref:#runtime[the runtime], * xref:#devenv[the development environment support], * xref:#sources[the sources], and * xref:#distribution[the distribution], then xref:#details[details] some of the language and compiler changes, and finally points readers to the bug database for any xref:#knownLimitations[known limitations]. ''''' [[language]] == The Language AspectJ 1.1 is a slightly different language than AspectJ 1.0. In all but a few cases, programs written in AspectJ 1.0 should compile correctly in AspectJ 1.1. In many cases, there are new or preferred forms in AspectJ 1.1. However, some AspectJ 1.0 features have changed in 1.1, so some 1.0 programs will not compile or will run differently in 1.1. The corresponding features are marked below as compile-time or run-time incompatible (_CTI_ or _RTI_, respectively). When the language change involves a move in the static shadow effective at run-time but also apparent at compile-time (e.g., in declare error or warning statements), it is marked _CRTI_. Programs using run-time incompatible forms should be verified that they are behaving as expected in 1.1. Most changes to the language are additions to expressibility requested by our users: * xref:#THROWS_PATTERN[Matching based on throws]: You can now make finer discriminations between methods based on declared exceptions. * xref:#NEW_PCDS[New kinded pointcut designators]: Now every kind of join point has a corresponding kinded pointcut designator. Some are have different behavior in edge cases but offer improved power and clarity: * xref:#ASPECT_PRECEDENCE[New aspect precedence form]: AspectJ 1.1 has a new declare form, `declare precedence`, that replaces the "dominates" clause on aspects. (_CTI_) * The order of xref:#SUPER_IFACE_INITS[initialization join points for super-interfaces] has been clarified. (_RTI_) But in order to support weaving into bytecode effectively, several incompatible changes had to be made to the language: * A class's default constructor may xref:#DEFAULT_CONSTRUCTOR_CONFLICT[conflict] with an inter-type constructor. (_CTI_) * xref:#NO_CALLEE_SIDE_CALL[No callee-side call join points]: The AspectJ 1.1 compiler does not expose call join points unless it is given the calling code. (_CRTI_) * xref:#SINGLE_INTERCLASS_TARGET[One target for intertype declarations]. (_CTI_) * xref:#UNAVAILABLE_JOIN_POINTS[No initializer execution join points]. (_RTI_) * xref:#AFTER_HANDLER[No after or around advice on handler join points]. (_CTI_) * xref:#CONSTRUCTOR_EXECUTION_IS_BIGGER[Initializers run inside constructor execution join points]. (_RTI_) * xref:#INTER_TYPE_FIELD_INITIALIZERS[inter-type field initializers] run before class-local field initializers. (_RTI_) * xref:#WITHIN_MEMBER_TYPES[Small limitations of the within pointcut.] (_CRTI_) * xref:#WITHIN_CODE[Small limitations of the withincode pointcut.] (_CRTI_) * xref:#INSTANCEOF_ON_WILD[Can't do instanceof matching on type patterns with wildcards]. (_CTI_) * xref:#NO_SOURCE_COLUMN[SourceLocation.getColumn() is deprecated and will always return 0]. (_RTI_) * The interaction between aspect instantiation and advice has been xref:#ASPECT_INSTANTIATION_AND_ADVICE[clarified]. (_RTI_) * xref:#STRINGBUFFER[The String + operator is now correctly advised]. (_CRTI_) [#NEW_LIMITATIONS]#There# are a couple of language limitations for things that are rarely used that make the implementation simpler, so we have restricted the language accordingly. * xref:#VOID_FIELD_SET[Field set join points now have a `void` return type.] This will require porting of code that uses the `set` PCD in conjunction with after-returning or around advice. (_CTI_) * 'declare soft: TYPE: POINTCUT;' - AspectJ 1.1 only accepts TYPE rather than a TYPE_PATTERN. This limitation makes declare soft much easier to implement efficiently. (_CTI_) * Inter-type field declarations only allow a single field per line, i.e. this is now illegal 'int C.field1, D.field2;' This must instead be, 'int C.field1; int D.field2;' (_CTI_) * We did not implement the handling of more than one `..` wildcard in args PCD's (rarely encountered in the wild) because we didn't have the time. This might be available in later releases if there is significant outcry. (_CTI_) We did not implement the long-awaited xref:#PER_TYPE[new pertype aspect specifier] in this release, but it may well be in a future release. ''''' [[compiler]] == The Compiler The compiler for AspectJ 1.1 is different than the compiler for AspectJ 1.0. While this document describes the differences in the compiler, it's worthwhile noting that much effort has been made to make sure that the interface to ajc 1.1 is, as much as possible, the same as the interface to ajc 1.0. There are two important changes under the hood, however. First, the 1.1 compiler is implemented on top of the open-source Eclipse compiler. This has two benefits: It allows us to concentrate on the AspectJ extensions to Java and let the Eclipse team worry about making sure the Java edge cases work, and it allows us to piggyback on Eclipse's already mature incremental compilation facilities. Second, ajc now cleanly delineates compilation of source code from assembly (or "weaving") of bytecode. The compiler still accepts source code, but internally it transforms it into bytecode format before weaving. This new architecture, and other changes to the compiler, allows us to implement some features that were defined in the AspectJ 1.0 language but not implementable in the 1.1 compiler. It also makes some new features available: * xref:#SOURCEROOT[The -sourceroots option] takes one or more directories, and indicates that all the source files in those directories should be passed to the compiler. * xref:#BYTECODE_WEAVING[The -injars option] takes one or more jar files, and indicates that all the classfiles in the jar files should be woven into. * xref:#BINARY_ASPECTS[The -aspectpath option] takes one or more jar files, and weaves any aspects in .class form into the sources. * xref:#OUTJAR[The -outjar option] indicates that the result classfiles of compiling and weaving should be placed in the specified jar file. * xref:#XLINT[The -Xlint option] allows control over warnings. * xref:#OTHER_X_OPTIONS[Various -X options] changed. * xref:#INCREMENTAL[The -incremental option] tells the AspectJ 1.1 compiler to recompile only as necessary. Some other features we wanted to support for 1.1, but did not make it into this release: * xref:#ERROR_MESSAGES[Error messages will sometimes be scary] * xref:#MESSAGE_CONTEXT[Source code context is not shown for errors and warnings detected during bytecode weaving] But some features of the 1.0 compiler are not supported in the 1.1 compiler: * xref:#NO_SOURCE[The source-related options] -preprocess, -usejavac, -nocomment and -workingdir * xref:#NO_STRICT_LENIENT[The -strict and -lenient options] * xref:#NO_PORTING[The -porting option] * xref:#_13_REQUIRED[J2SE 1.2 is not supported; J2SE 1.3 or later is required.] A short description of the options ajc accepts is available with "`ajc -help`". Longer descriptions are available in the link:devguide/ajc-ref.html[Development Environment Guide section on ajc]. Some changes to the implementation are almost entirely internal: * The behavior of the compiler in xref:#TARGET_TYPES_MADE_PUBLIC[lifting the visibility] of the target types of some declares and pointcuts to public has been clarified. Also, it is worth noting that because AspectJ now works on bytecode, it is somewhat sensitive to how different compilers generate bytecode, especially when compiling with and without xref:#ONE_FOUR_METHOD_SIGNATURES[the -1.4 flag]. ''''' [[tools]] == Support Tools This release includes an Ant task for old-style 1.0 build scripts, a new task for all the new compiler options, and a CompilerAdapter to support running `ajc` with the Javac task by setting the `build.compiler` property. The new task can automatically copy input resources to output and work in incremental mode using a "tag" file. This release does not include `ajdoc`, the documentation tool for AspectJ sources. Ajdoc is deeply dependent on the abstract syntax tree classes from the old compiler, so it needs a bottom-up rewrite. We think it best to use this opportunity to implement more general API's for publishing and rendering static structure. Because those API's are last to settle in the new architecture, and because the compiler itself is a higher priority, we are delaying work on ajdoc until after the 1.1 release. AspectJ 1.1 will not include ajdb, the AspectJ stand-alone debugger. It is no longer necessary for two reasons. First, the -XnoInline flag will tell the compiler to generate code without inlining that should work correctly with any Java debugger. For code generated with inlining enabled, more third-party debuggers are starting to work according to JSR 45, "Debugging support for other languages," which is supported by AspectJ 1.0. We aim to support JSR-45 in AspectJ 1.1, but support will not be in the initial release. Consider using the -XnoInline flag until support is available. ''''' [[runtime]] == The Runtime Library This release has minor additions to the runtime library classes. As with any release, you should compile and run with the runtime library that came with your compiler, and you may run with a later version of the library without recompiling your code. In one instance, however, runtime classes behave differently this release. Because the AspectJ 1.1 compiler does its weaving through bytecode, column numbers of source locations are not available. Therefore, `thisJoinPoint.getSourceLocation().getColumn()` is deprecated and will always return 0. ''''' [[devenv]] == The AJDE Tools The AspectJ Browser supports incremental compilation and running programs. AJDE for JBuilder, AJDE for NetBeans, and AJDE for Emacs are now independent SourceForge projects (to keep their licenses). They use the batch-build mode of the new compiler. ''''' [[sources]] == The Sources and the Licenses The AspectJ tools sources are available under the https://www.eclipse.org/org/documents/epl-2.0/EPL-2.0.txt[Eclipse Public License v 2.0] in the Git repository at https://eclipse.org/aspectj. For more information, see the FAQ entry on building sources. ''''' [[distribution]] == The AspectJ distribution AspectJ 1.0 had many distributions - for the tools, the documentation, each IDE support package, their respective sources, and the Ant tasks - because they came under different licenses. All of AspectJ 1.1 is licensed under the CPL 1.0, so the tools, Ant tasks, and documentation are all in one distribution available from https://eclipse.org/aspectj. To retain their MPL 1.1 license, Ajde for http://aspectj4emacs.sourceforge.net/[Emacs], http://aspectj4netbean.sourceforge.net/[NetBeans] and http://aspectj4jbuildr.sourceforge.net/[JBuilder] are now independent SourceForge projects. ''''' ''''' [[details]] == Details of some language and compiler changes [[ASPECT_INSTANTIATION_AND_ADVICE]] === Aspect Instantiation and Advice In AspectJ 1.0.6, we made an effort to hide some complications with Aspect instantiation from the user. In particular, the following code compiled and ran: [source, java] .... public class Client { public static void main(String[] args) { Client c = new Client(); } } aspect Watchcall { pointcut myConstructor(): execution(new(..)); before(): myConstructor() { System.err.println("Entering Constructor"); } } .... But there's a conceptual problem with this code: The before advice should run before the execution of all constructors in the system. It must run in the context of an instance of the Watchcall aspect. The only way to get such an instance is to have Watchcall's default constructor execute. But before that executes, we need to run the before advice... AspectJ 1.0.6 hid this circularity through the ad-hoc mechanism of preventing an aspect's advice from matching join points that were within the aspect's definition, and occurred before the aspect was initialized. But even in AspectJ 1.0.6, this circularity could be exposed: [source, java] .... public class Client { public static int foo() { return 3; } public static void main(String[] args) { Client c = new Client(); } } aspect Watchcall { int i = Client.foo(); pointcut myConstructor(): execution(new(..)) || execution(int foo()); before(): myConstructor() { System.err.println("Entering Constructor"); } } .... This program would throw a NullPointerException when run, since Client.foo() was called before the Watchcall instance could be instantiated. In AspectJ 1.1, we have decided that half-hiding the problem just leads to trouble, and so we are no longer silently hiding some join points before aspect initialization. However, we have provided a better exception than a NullPointerException for this case. In AspectJ 1.1, both of the above programs will throw org.aspectj.lang.NoAspectBoundException. [[THROWS_PATTERN]] === Matching based on throws Type patterns may now be used to pick out methods and constructors based on their throws clauses. This allows the following two kinds of extremely wildcarded pointcuts: [source, java] .... pointcut throwsMathlike(): // each call to a method with a throws clause containing at least // one exception with "Math" in its name. call(* *(..) throws *..*Math*); pointcut doesNotThrowMathlike(): // each call to a method with a throws clause containing no // exceptions with "Math" in its name. call(* *(..) throws !*..*Math*); .... The longwinded rules are that a method or constructor pattern can have a "throws clause pattern". Throws clause patterns look like: [source, text] .... ThrowsClausePattern: ThrowsClausePatternItem ("," ThrowsClausePatternItem)* ThrowsClausePatternItem: ["!"] TypeNamePattern .... A ThrowsClausePattern matches the ThrowsClause of any code member signature. To match, each ThrowsClausePatternItem must match the throws clause of the member in question. If any item doesn't match, then the whole pattern doesn't match. This rule is unchanged from AspectJ 1.0. If a ThrowsClausePatternItem begins with "!", then it matches a particular throws clause if and only if _none_ of the types named in the throws clause is matched by the TypeNamePattern. If a ThrowsClausePatternItem does not begin with "!", then it matches a throws clause if and only if _any_ of the types named in the throws clause is matched by the TypeNamePattern. These rules are completely backwards compatible with AspectJ 1.0. The rule for "!" matching has one potentially surprising property, in that the two PCD's shown below will have different matching rules. [source, java] .... /*[1]*/ call(* *(..) throws !IOException) /*[2]*/ call(* *(..) throws (!IOException)) void m() throws RuntimeException, IOException {} .... [1] will NOT match the method m(), because method m's throws clause declares that it throws IOException. [2] WILL match the method m(), because method m's throws clause declares the it throws some exception which does not match IOException, i.e. RuntimeException. [[NEW_PCDS]] === New kinded pointcut designators AspectJ 1.0 does not provide kinded pointcut designators for two (rarely used) join points: preinitialization (the code that runs before a super constructor call is made) and advice execution. AspectJ 1.1 does not change the meaning of the join points, but provides two new pointcut designators to pick out these join points, thus making join points and pointcut designators more parallel. `adviceexectuion()` will pick out advice execution join points. You will usually want to use `adviceexecution() && within(Aspect)` to restrict it to only those pieces of advice defined in a particular aspect. + `preinitialization(ConstructorPattern)` will pick out pre-initialization join points where the initialization process is entered through `ConstructorPattern`. [[PER_TYPE]] === New pertype aspect specifier (not in 1.1) We strongly considered adding a pertype aspect kind to 1.1. This is somewhat motivated by the new xref:#SINGLE_INTERCLASS_TARGET[restrictions on inter-type declarations] . This is also motivated by many previous request to support a common logging idiom. Here's what pertype would look like: [source, java] .... /** One instance of this aspect will be created for each class, * interface or aspect in the com.bigboxco packages. */ aspect Logger pertype(com.bigboxco..*) { /* This field holds a logger for the class. */ Log log; /* This advice will run for every public execution defined by * a type for which a Logger aspect has been created, i.e. * any type in com.bigboxco..* */ before(): execution(public * *(..)) { log.enterMethod(thisJoinPoint.getSignature().getName()); } /* We can use a special constructor to initialize the log field */ public Logger(Class myType) { this.log = new Log(myType); } } /** External code could use aspectOf to get at the log, i.e. */ Log l = Logger.aspectOf(com.bigboxco.Foo.class).log; .... The one open question that we see is how this should interact with inner types. If a pertype aspect is created for an outer type should advice in that aspect run for join points in inner types? That is the behavior of the most common uses of this idiom. In any case, this feature will not be in AspectJ 1.1. [[SINGLE_INTERCLASS_TARGET]] === One target for intertype declarations Intertype declarations (once called "introductions") in AspectJ 1.1 can only have one target type. So the following code intended to declare that there is a void doStuff() method on all subtypes of Target is not legal AspectJ 1.1 code. [source, java] .... aspect A { public void Target+.doStuff() { ... } } .... The functionality of "multi-intertype declarations" can be recovered by using a helper interface. [source, java] .... aspect A { private interface MyTarget {} declare parents: Target+ implements MyTarget; public void MyTarget.doStuff() { ... } } .... We believe this is better style in AspectJ 1.0 as well, as it makes clear the static type of "this" inside the method body. The one piece of functionality that can not be easily recovered is the ability to add static fields to many classes. We believe that the xref:#PER_TYPE[pertype proposal] provides this functionality in a much more usable form. [[UNAVAILABLE_JOIN_POINTS]] === No initializer execution join points AspectJ 1.1 does not consider initializer execution a principled join point. The collection of initializer code (the code that sets fields with initializers and the code in non-static initializer blocks) is something that makes sense only in Java source code, not in Java bytecode. [[AFTER_HANDLER]] === No after or around advice on handler join points The end of an exception handler is underdetermined in bytecode, so ajc will not implement after or around advice on handler join points, instead signaling a compile-time error. [[CONSTRUCTOR_EXECUTION_IS_BIGGER]] === Initializers run inside constructor execution join points The code generated by the initializers in Java source code now runs inside of constructor execution join points. This changes how before advice runs on constructor execution join points. Consider: [source, java] .... class C { C() { } String id = "identifier"; // this assignment // has to happen sometime } aspect A { before(C c) this(c) && execution(C.new()) { System.out.println(c.id.length()); } } .... In AspectJ 1.0, this will print "10", since id is assigned its initial value prior to the before advice's execution. However, in AspectJ 1.1, this will throw a NullPointerExcception, since "id" does not have a value prior to the before advice's execution. Note that the various flavors of after returning advice are unchanged in this respect in AspectJ 1.1. Also note that this only matters for the execution of constructors that call a super-constructor. Execution of constructors that call a this-constructor are the same in AspectJ 1.1 as in AspectJ 1.0. We believe this difference should be minimal to real programs, since programmers using before advice on constructor execution must always assume incomplete object initialization, since the constructor has not yet run. [[INTER_TYPE_FIELD_INITIALIZERS]] === Inter-type field initializers The initializer, if any, of an inter-type field definition runs before the class-local initializers of its target class. In AspectJ 1.0.6, such an initializer would run after the initializers of a class but before the execution of any of its constructor bodies. As already discussed in the sections about xref:#UNAVAILABLE_JOIN_POINTS[initializer execution join points] and xref:#CONSTRUCTOR_EXECUTION_IS_BIGGER[constructor execution], the point in code between the initializers of a class and its constructor body is not principled in bytecode. So we had a choice of running the initializer of an inter-type field definition at the beginning of initialization (i.e., before initializers from the target class) or at the end (i.e., just before its called constructor exits). We chose the former, having this pattern in mind: [source, java] .... int C.methodCount = 0; before(C c): this(c) && execution(* *(..)) { c.methodCount++; } .... We felt there would be too much surprise if a constructor called a method (thus incrementing the method count) and then the field was reset to zero after the constructor was done. [[WITHIN_MEMBER_TYPES]] === Small limitations of the within pointcut Because of the guarantees made (and not made) by the Java classfile format, there are cases where AspectJ 1.1 cannot guarantee that the within pointcut designator will pick out all code that was originally within the source code of a certain type. The non-guarantee applies to code inside of anonymous and local types inside member types. While the within pointcut designator behaves exactly as it did in AspectJ 1.0 when given a package-level type (like C, below), if given a member-type (like C.InsideC, below), it is not guaranteed to capture code in contained local and anonymous types. For example: [source, java] .... class C { Thread t; class InsideC { void setupOuterThread() { t = new Thread( new Runnable() { public void run() { // join points with code here // might not be captured by // within(C.InsideC), but are // captured by within(C) System.out.println("hi"); } }); } } } .... We believe the non-guarantee is small, and we haven't verified that it is a problem in practice. [[WITHIN_CODE]] === Small limitations of the withincode pointcut The withincode pointcut has similar issues to those described above for within. [[INSTANCEOF_ON_WILD]] === Can't do instanceof matching on type patterns with wildcard The pointcut designators this, target and args specify a dynamic test on their argument. These tests can not be performed on type patterns with wildcards in them. The following code that compiled under 1.0 will be an error in AspectJ-1.1: [source, java] .... pointcut oneOfMine(): this(com.bigboxco..*); .... The only way to implement this kind of matching in a modular way would be to use the reflection API at runtime on the Class of the object. This would have a very high performance cost and possible security issues. There are two good work-arounds. If you control the source or bytecode to the type you want to match then you can use declare parents, i.e.: [source, java] .... private interface OneOfMine {} declare parents: com.bigboxco..* implements OneOfMine; pointcut oneOfMine(): this(OneOfMine); .... If you want the more dynamic matching and are willing to pay for the performance, then you should use the Java reflection API combined with if. That would look something like: [source, java] .... pointcut oneOfMine(): this(Object) && if(classMatches("com.bigboxco..*", thisJoinPoint.getTarget().getClass())); static boolean classMatches(String pattern, Class _class) { if (patternMatches(pattern, _class.getName())) return true; ... } .... Note: wildcard type matching still works in all other PCD's that match based on static types. So, you can use 'within(com.bigboxco..*+)' to match any code lexically within one of your classes or a subtype thereof. This is often a good choice. [[NO_SOURCE_COLUMN]] === SourceLocation.getColumn() The Java .class file format contains information about the source file and line numbers of its contents; however, it has no information about source columns. As a result, we can not effectively support the access of column information in the reflection API. So, any calls to thisJoinPoint.getSourceLocation().getColumn() will be marked as deprecated by the compiler, and will always return 0. [[ASPECT_PRECEDENCE]] === Aspect precedence AspectJ 1.1 has a new declare form: [source, java] .... declare precedence ":" TypePatternList ";" .... This is used to declare advice ordering constraints on join points. For example, the constraints that (1) aspects that have Security as part of their name should dominate all other aspects, and (2) the Logging aspect (and any aspect that extends it) should dominate all non-security aspects, can be expressed by: [source, java] .... declare precedence: *..*Security*, Logging+, *; .... In the TypePatternList, the wildcard * means "any type not matched by another type in the declare precedence". ==== Various cycles It is an error for any aspect to be matched by more than one TypePattern in a single declare precedence, so: [source, java] .... declare precedence: A, B, A ; // error .... However, multiple declare precedence forms may legally have this kind of circularity. For example, each of these declare precedence is perfectly legal: [source, java] .... declare precedence: B, A; declare precedence: A, B; .... And a system in which both constraints are active may also be legal, so long as advice from A and B don't share a join point. So this is an idiom that can be used to enforce that A and B are strongly independent. ==== Applies to concrete aspects Consider the following library aspects: [source, java] .... abstract aspect Logging { abstract pointcut logged(); before(): logged() { System.err.println("thisJoinPoint: " + thisJoinPoint); } } aspect MyProfiling { abstract pointcut profiled(); Object around(): profiled() { long beforeTime = System.currentTimeMillis(); try { return proceed(); } finally { long afterTime = System.currentTimeMillis(); addToProfile(thisJoinPointStaticPart, afterTime - beforeTime); } } abstract void addToProfile( org.aspectj.JoinPoint.StaticPart jp, long elapsed); } .... In order to use either aspect, they must be extended with concrete aspects, say, MyLogging and MyProfiling. In AspectJ 1.0, it was not possible to express that Logging's advice (when concerned with the concrete aspect MyLogging) dominated Profiling's advice (when concerned with the concrete aspect MyProfiling) without adding a dominates clause to Logging itself. In AspectJ 1.1, we can express that constraint with a simple: [source, java] .... declare precedence: MyLogging, MyProfiling; .... ==== Changing order of advice for sub-aspects By default, advice in a sub-aspect has more precedence than advice in a super-aspect. One use of the AspectJ 1.0 dominates form was to change this precedence: [source, java] .... abstract aspect SuperA dominates SubA { pointcut foo(): ... ; before(): foo() { // in AspectJ 1.0, runs before the advice in SubA // because of the dominates clause } } aspect SubA extends SuperA { before(): foo() { // in AspectJ 1.0, runs after the advice in SuperA // because of the dominates clause } } .... This no longer works in AspectJ 1.1, since declare precedence only matters for concrete aspects. Thus, if you want to regain this kind of precedence change, you will need to refactor your aspects. [[SOURCEROOT]] === The -sourceroots option The AspectJ 1.1 compiler now accepts a -sourceroots option used to pass all .java files in particular directories to the compiler. It takes either a single directory name, or a list of directory names separated with the CLASSPATH separator character (":" for various Unices, ";" for various Windows). So, if you have your project separated into a gui module and a base module, each of which is stored in a directory tree, you might use one of [source, text] .... ajc -sourceroots /myProject/gui:/myProject/base ajc -sourceroots d:\myProject\gui;d:\myProject\base .... This option may be used in conjunction with lst files, listing .java files on the command line, and the -injars option. [[BYTECODE_WEAVING]] === The -injars option The AspectJ 1.1 compiler now accepts an -injars option used to pass all .class files in a particular jar file to the compiler. It takes either a single directory name, or a list of directory names separated with the CLASSPATH separator character (":" for various Unices, ";" for various Windows). So, if MyTracing.java defines a trace aspect that you want to apply to all the classes in myBase.jar and myGui.jar, you would use one of: [source, text] .... ajc -injars /bin/myBase.jar:/bin/myGui.jar MyTracing.java ajc -injars d:\bin\myBase.jar;d:\bin\myGui.jar MyTracing.java .... The class files in the input jars must not have had advice woven into them, since AspectJ enforces the requirement that advice is woven into a particular classfile only once. So if the classfiles in the jar file are to be created with the ajc compiler (as opposed to a pure Java compiler), they should not be compiled with any non-abstract aspects. This option may be used in conjunction with lst files, listing .java files on the command line, and the -sourceroots option. [[OUTJAR]] === The -outjar option The -outjar option takes the name of a jar file into which the results of the compilation should be put. For example: [source, text] .... ajc -injars myBase.jar MyTracing.java -outjar myTracedBase.jar .... No meta information is placed in the output jar file. [[INCREMENTAL]] === Incremental compilation The AspectJ 1.1 compiler now supports incremental compilation. When ajc is called with the -incremental option, it must also be passed a -sourceroots option specifying a directory to incrementally compile. Once the initial compile is done, ajc waits for console input. Every time it reads a new line (i.e., every time the user hits return) ajc recompiles those input files that need recompiling. ==== Limitations This new functionality is still only lightly tested. [[XNOWEAVE]] === -XnoWeave, a compiler option to suppress weaving The -XnoWeave option suppresses weaving, and generates classfiles and that can be passed to ajc again (through the -injars option) to generate final, woven classfiles. This option was originally envisioned to be the primary way to generate binary aspects that could be linked with other code, and so it was previously (in AspectJ 1.1beta1) named `-noweave`. We feel that using the `-aspectpath` option is a much better option. There may still be use cases for unwoven classfiles, but we've moved the flag to experimental status. [[BINARY_ASPECTS]] === -aspectpath, working with aspects in .class/.jar form When aspects are compiled into classfiles, they include all information necessary for the ajc compiler to weave their advice and deal with their inter-type declarations. In order for these aspects to have an effect on a compilation process, they must be passed to the compiler on the -aspectpath. Every .jar file on this path will be searched for aspects and any aspects that are found will be enabled during the compilation. The binary forms of this aspects will be untouched. [[NO_CALLEE_SIDE_CALL]] === Callee-side call join points The 1.0 implementation of AspectJ, when given: [source, java] .... class MyRunnable implements Runnable { public void run() { /*...*/ } } aspect A { call(): (void run()) && target(MyRunnable) { // do something here } } .... would cause A's advice to execute even when, say, java.lang.Thread called run() on a MyRunnable instance. With the new compiler, two things have happened in regard to callee-side calls: . because the programmer has access to more code (i.e., bytecode, not just source code), callee-side calls are much less important to have. . because compilation is more modular, allowing and encouraging separate compilation, callee-side calls are much more difficult to implement With these two points in mind, advice in an aspect will not be applied to call join points whose call site is completely unavailable to the aspect. . One reason (though not the only reason) we worked so hard in the _implementation_ of 1.0.6 to expose call join points, even if we only had access to the callee's code, was that otherwise users couldn't get access to call join points where the call was made from bytecode. This is no longer the case. In short, the implementation controls much more code (or has the capability to) than ever before. . The implementation model for the AspectJ 1.1 compiler is to separate the compilation of aspects/advice from their weaving/linking. A property of the model is that the compilation requires no access to "target" code, only the weaving/linking does, and weaving/linking is inherently per-class local: No action at weaving/linking time depends on the coordinated mangling of multiple classfiles. Rather, all weaving is done on a per classfile basis. This is an essential property for the current separate compilation model. + However, allowing implementation of call advice on either side requires simultaneous knowledge of both sides. If we first have access to a call, we can't decide to simply put the advice on the call site, since later we may decide to implement on the callee. This implementation decision is completely in the letter and the spirit of the AspectJ language. From the semantics guide describing code the implementation controls: ____ But AspectJ implementations are permitted to deviate from this in a well-defined way -- they are permitted to advise only accesses in _code the implementation controls_. Each implementation is free within certain bounds to provide its own definition of what it means to control code. ____ And about a particular decision about the 1.0.6 implementation: ____ Different join points have different requirements. Method call join points can be advised only if ajc controls _either_ the code for the caller or the code for the receiver, and some call pointcut designators may require caller context (what the static type of the receiver is, for example) to pick out join points. ____ The 1.1 implementation makes a different design decision: Method call join points can be advised only if ajc (in compiler or linker form) controls the code for the caller. What does 1.1 gain from this? * a clear (and implemented) separate compilation model (see point 2, above) * a less confusing interaction between call join points and the thisJoinPoint reflective object: We still get bug reports about source information sometimes existing and sometimes not existing at call join points. What does 1.1 lose from this? * The ability to capture all calls to Runnable.run() from anywhere to code ajc has access too, even from Thread, even if you don't compile java.lang with ajc. * The ability to, without access to the caller, capture entry to a particular method, but not super calls. * A code-size-improvement performance optimization. What are the possibilities for the future? * AspectJ 1.1.1 could expand its capture of call join points, possibly at the expense of separate compilation clarity, possibly not. * AspectJ 1.1.1 could re-introduce reception join points from AspectJ 0.7 (what callee-side call join points actually are): though they would never ever be taught in a tutorial or entry-level description of the model, they may have specialized uses. How will this affect developers? * When using the call PCD but only supplying the callee code, supply the calling code or use the execution PCD instead. [[OTHER_X_OPTIONS]] === Various -X options The AspectJ 1.0 compiler supported a number of options that started with X, for "experimental". Some of them will not be supported in 1.1, either because the "experiment" succeeded (in which case it's part of the normal functionality) or failed. Others will be supported as is (or nearly so) in 1.1: * -XOcodeSize: This is no longer necessary because inlining of around advice is on by default. We support its inverse, xref:#XNOINLINE[`-XnoInline`]. * xref:#XNOWEAVE[-XnoWeave, a compiler option to suppress weaving] * -XtargetNearSource: Not supported in this release. * -XserializableAspects: Supported. * -XaddSafePrefix: This option will not be supported in 1.1 at all because we're now always using (what we believe to be) safe prefixes. * -Xlint: Still supported, with xref:#XLINT[various options]. [[ERROR_MESSAGES]] === Some confusing error messages Building on the eclipse compiler has given us access to a very sophisticated problem reporting system as well as highly optimized error messages for pure Java code. Often this leads to noticeably better error messages than from ajc-1.0.6. However, when we don't handle errors correctly this can sometimes lead to cascading error messages where a single small syntax error will produce dozens of other messages. Please report any very confusing error messages as bugs. [[MESSAGE_CONTEXT]] === Source code context is not shown for errors and warnings detected during bytecode weaving For compiler errors and warnings detected during bytecode weaving, source code context will not be displayed. In particular, for declare error and declare warning statements, the compiler now only emits the file and line. We are investigating ways to overcome this in cases where the source code is available; in cases where source code is not available, we might specify the signature of the offending code. For more information, see bug 31724. [[XLINT]] === The -Xlint option `-Xlint:ignore,error,warning` will set the level for all Xlint warnings. `-Xlint`, alone, is an abbreviation for `-Xlint:warning`. The `-Xlintfile:lint.properties` allows fine-grained control. In tools.jar, see `org/aspectj/weaver/XlintDefault.properties` for the default behavior and a template to copy. More `-Xlint` warnings are supported now, and we may add disabled warnings in subsequent bug-fix releases of 1.1. Because the configurability allows users to turn off warnings, we will be able to warn about more potentially dangerous situations, such as the potentially unsafe casts used by very polymorphic uses of proceed in around advice. [[NO_SOURCE]] === Source-specific options Because AspectJ 1.1 does not generate source code after weaving, the source-code-specific options -preprocess, -usejavac, -nocomment and -workingdir options are meaningless and so not supported. [[NO_STRICT_LENIENT]] === The -strict and -lenient options Because AspectJ 1.1 uses the Eclipse compiler, which has its own mechanism for changing strictness, we no longer support the -strict and -lenient options. [[NO_PORTING]] === The -porting option AspectJ 1.1 does not have a -porting option. [[_13_REQUIRED]] === J2SE 1.3 required Because we build on Eclipse, the compiler will no longer run under J2SE 1.2. You must run the compiler (and all tools based on the compiler) using J2SE 1.3 or later. The code generated by the compiler can still run on Java 1.1 or later VM's if compiled against the correct runtime libraries. [[DEFAULT_CONSTRUCTOR_CONFLICT]] === Default constructors AspectJ 1.1 does not allow the inter-type definition of a zero-argument constructor on a class with a visible default constructor. So this is no longer allowed: [source, java] .... class C {} aspect A { C.new() {} // was allowed in 1.0.6 // is a "multiple definitions" conflict in 1.1 } .... In the Java Programming Language, a class defined without a constructor actually has a "default" constructor that takes no arguments and just calls `super()`. This default constructor is a member of the class like any other member, and can be referenced by other classes, and has code generated for it in classfiles. Therefore, it was an oversight that AspectJ 1.0.6 allowed such an "overriding" inter-type constructor definition. [[SUPER_IFACE_INITS]] === Initialization join points for super-interfaces In AspectJ, interfaces may have non-static members due to inter-type declarations. Because of this, the semantics of AspectJ defines the order that initializer code for interfaces is run. In the semantics document for AspectJ 1.0.6, the following promises were made about the order of this initialization: . a supertype is initialized before a subtype . initialized code runs only once . initializers for supertypes run in left-to-right order The first two properties are important and are preserved in AspectJ 1.1, but the third property is and was ludicrous, and was never properly implemented (and never could be) in AspectJ 1.0.6. Consider: [source, java] .... interface Top0 {} interface Top1 {} interface I extends Top0, Top1 {} interface J extends Top1, Top0 {} class C implements I, J {} // I says Top0's inits must run before Top1's // J says Top1's inits must run before Top0's aspect A { int Top0.i = foo("I'm in Top0"); int Top1.i = foo("I'm in Top1"); static int foo(String s) { System.out.println(s); return 37; } } .... This was simply a bug in the AspectJ specification. The correct third rule is: ____ the initializers for a type's superclass are run before the initializers for its superinterfaces. ____ [[VOID_FIELD_SET]] === Field Set Join Points In AspectJ 1.0.6, the join point for setting a field F had, as a return type, F's type. This was "java compatible" because field assignment in java, such as "Foo.i = 37", is in fact an expression, and does in fact return a value, the value that the field is assigned to. This was never "java programmer compatible", however, largely because programmers have absorbed the good style of rarely using an assignment statement in a value context. Programmers typically expect "Foo.i = 37" not to return a value, but to simply assign a value. Thus, programmers typically wanted to write something like: [source, java] .... void around(): set(int Foo.i) { if (theSetIsAllowed()) { proceed(); } } .... And were confused by it being a compile-time error. They weren't confused for long, and soon adapted to writing: [source, java] .... int around(): set(int Foo.i) { if (theSetIsAllowed()) { return proceed(); } else { return Foo.i; } } .... But there was definitely a short disconnect. On top of that, we were never shown a convincing use-case for returning an interesting value from a set join point. When we revisited this issue, in fact, we realized we had a long-standing bug in 1.0.6 dealing with the return value of pre-increment expressions (such as ++Foo.i) that nobody had found because nobody cares about the return value of such join points. So, because it's easier to implement, and because we believe that this is the last possibility to make the semantics more useful, we have made set join points have a void return type in 1.1. [[XNOINLINE]] === The -XnoInline Option The `-XnoInline` option to indicate that no inlining of any kind should be done. This is purely a compiler pragma: No program semantics (apart from stack traces) will be changed by the presence or absence of this option. [[TARGET_TYPES_MADE_PUBLIC]] === Target types made public Even in 1.0.6, the AspectJ compiler has occasionally needed to convert the visibility of a package-level class to a public one. This was previously done in an ad-hoc basis that took whole-program analysis into account. With the incremental compilation model of AspectJ 1.1, we can now specify the occasions when the compiler makes these visibility changes. In particular, the types used in the `this`, `target`, and `args` pointcuts are made public, as are the super-types from `declare parents` and the exception type from `declare soft`. We believe the visibility changes could be avoided in the future with various implementation tricks if they become a serious concern, but did not encounter them as such a concern when they were done in the 1.0.6 implementation. [[STRINGBUFFER]] === String + now advised In Java, the + operator sometimes results in StringBuffer objects being created, appended to, and used to generate a new String. Thus, [source, java] .... class Foo { String makeEmphatic(String s) { return s + "!"; } } .... is approximately the same at runtime as [source, java] .... class Foo { String makeEmphatic(String s) { return new StringBuffer(s).append("!").toString(); } } .... In the design process of AspectJ 1.0.6 we didn't expose those StringBuffer methods and constructors as join points (though we did discuss it), but in 1.1 we do. This change is likely to affect highly wildcarded aspects, and can do so in surprising ways. In particular: [source, java] .... class A { before(int i): call(* *(int)) && args(i) { System.err.println("entering with " + i); } } .... may result in a stack overflow error, since the argument to println is really [source, java] .... new StringBuffer("entering with ").append(i).toString() .... which has a call to StringBuffer.append(int). In such cases, it's worth restricting your pointcut, with something like one of: [source, java] .... call(* *(int)) && args(i) && !within(A) call(* *(int)) && args(i) && !target(StringBuffer) .... [[ONE_FOUR_METHOD_SIGNATURES]] === The -1.4 flag and method signatures Consider the following aspect [source, java] .... public aspect SwingCalls { pointcut callingAnySwing(): call(* javax.swing..*+.*(..)); before(): callingAnySwing() { System.out.println("Calling any Swing"); } } .... And then consider the two statements [source, java] .... JFrame frame = new JFrame(); frame.setTitle("Title"); .... According to the Java Language Specification version 2, the call to `frame.setTitle("Title")` should always produce the bytecode for a call to `javax.swing.JFrame.setTitle`. However, older compilers (and eclipse when run without the `-1.4` flag) will generate the bytecode for a call to `java.awt.Frame.setTitle` instead since this method is not overriden by JFrame. The AspectJ weaver depends on the correctly generated bytecode in order to match patterns like the one you show correctly. This is a good example of why the pattern `call(* *(..)) && target(JFrame)` is the recommended style. In general, OO programmers don't want to care about the static type of an object at a call site, but only want to know the dynamic instanceof behavior which is what the target matching will handle. [[knownLimitations]] == Known limitations The AspectJ 1.1.0 release contains a small number of known limitations relative to the AspectJ 1.1 language. For the most up-to-date information about known limitations in an AspectJ 1.1 release, see the bug database at https://bugs.eclipse.org/bugs, especially the open bugs for the https://bugs.eclipse.org/bugs/buglist.cgi?product=AspectJ&component=Compiler&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED[compiler], https://bugs.eclipse.org/bugs/buglist.cgi?product=AspectJ&component=IDE&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED[IDE support], https://bugs.eclipse.org/bugs/buglist.cgi?product=AspectJ&component=Doc&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED[documentation], and https://bugs.eclipse.org/bugs/buglist.cgi?product=AspectJ&component=Ant&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED[Ant tasks]. Developers should know about bugs marked with the "info" keyword because those bugs reflect failures to implement the 1.1 language perfectly. These might be fixed during the 1.1 release cycle; find them using the query https://bugs.eclipse.org/bugs/buglist.cgi?product=AspectJ&keywords=info For ajc's 1.1 implementation limitations, see link:progguide/implementation.html[Programming Guide Appendix: "Implementation Notes"].