]>
source.dussan.org Git - aspectj.git/log
aclement [Tue, 26 Jul 2005 15:10:20 +0000 (15:10 +0000)]
add the version when we add either aspect/pointcut attributes.
aclement [Tue, 26 Jul 2005 14:17:04 +0000 (14:17 +0000)]
sick sick sick generic itd tests.
aclement [Tue, 26 Jul 2005 14:15:32 +0000 (14:15 +0000)]
minor versioning improvement. dont add the version attribute if someone else already has.
aclement [Tue, 26 Jul 2005 14:01:14 +0000 (14:01 +0000)]
fixing the versioning problem. Up until now if we found no version in a class we 'assumed' it was the latest (1.5.0) - this is dumb. No version really means it is probably 1.2.0 (as the version was added for 1.2.1). These fixes mean that you can now safely check that version when reading in attributes, confident that checking for 1.5.0 *means* 1.5.0 was used to build it.
aclement [Tue, 26 Jul 2005 08:38:34 +0000 (08:38 +0000)]
generics: testcase - static generic method ITD
aclement [Tue, 26 Jul 2005 08:25:05 +0000 (08:25 +0000)]
generics: testcase - static generic method ITD
acolyer [Sat, 23 Jul 2005 07:10:18 +0000 (07:10 +0000)]
This commit simply makes the World a better place.
I've refactored, tidied, and commented. Members now have minimum necessary visibility (e.g fields are private). Some redundant methods are eliminated altogether. Also caches precedence calculations, but other than that there are no functional changes.
acolyer [Sat, 23 Jul 2005 07:08:10 +0000 (07:08 +0000)]
removed erroneous "s" from method name
acolyer [Sat, 23 Jul 2005 07:07:16 +0000 (07:07 +0000)]
added extra utility method for warnings
acolyer [Fri, 22 Jul 2005 15:38:36 +0000 (15:38 +0000)]
update to expected messages for type pattern matching with generics
acolyer [Fri, 22 Jul 2005 15:38:00 +0000 (15:38 +0000)]
annotated lines that should produce compilation warnings
acolyer [Fri, 22 Jul 2005 15:37:06 +0000 (15:37 +0000)]
also support superinterfaces when converting an Eclipse TypeParameter into a TypeVariable
acolyer [Fri, 22 Jul 2005 15:36:32 +0000 (15:36 +0000)]
improved matching in canBeBoundTo to consider equality of all bounds when testing against a type variable reference
acolyer [Fri, 22 Jul 2005 15:35:54 +0000 (15:35 +0000)]
remember source location when faulting in generic type
acolyer [Fri, 22 Jul 2005 15:34:14 +0000 (15:34 +0000)]
changed the serialized form of a WildTypePattern to remember generic info... standby for mass updates of binary comparison tests.
avasseur [Fri, 22 Jul 2005 14:57:40 +0000 (14:57 +0000)]
fix #104212 very bad bug on static method call jp that has been around since java 1.4 is there
acolyer [Thu, 21 Jul 2005 13:27:37 +0000 (13:27 +0000)]
proper support for type variable across patterns in the Eclipse Source Type world. Still need serialization support for this.
acolyer [Thu, 21 Jul 2005 13:26:28 +0000 (13:26 +0000)]
new warnings for generics in patterns
acolyer [Thu, 21 Jul 2005 13:26:10 +0000 (13:26 +0000)]
change from isTypeVariable to isTypeVariableReference
acolyer [Thu, 21 Jul 2005 13:25:36 +0000 (13:25 +0000)]
more tests and refinement of existing tests
acolyer [Thu, 21 Jul 2005 13:25:14 +0000 (13:25 +0000)]
fixes bug in test case caught by improved error messages in compiler
aclement [Thu, 21 Jul 2005 10:13:00 +0000 (10:13 +0000)]
dont go looking for generic types pre 1.5
aclement [Thu, 21 Jul 2005 10:01:29 +0000 (10:01 +0000)]
generics: fixes for decp testcases involving generic types targeted by parameterized decps. Type variables now persist in the system for longer... also LazyClassGen generates correct generic signatures for modified types.
aclement [Thu, 21 Jul 2005 08:25:23 +0000 (08:25 +0000)]
tidy up after a run...
avasseur [Thu, 21 Jul 2005 08:13:24 +0000 (08:13 +0000)]
fix in ltw log + test + some doc on custom message handler is evil
acolyer [Wed, 20 Jul 2005 16:18:53 +0000 (16:18 +0000)]
one more passing test...
acolyer [Wed, 20 Jul 2005 16:18:21 +0000 (16:18 +0000)]
wild type patterns with no type params ignore <...> when matching by name (raw match)
acolyer [Wed, 20 Jul 2005 16:17:52 +0000 (16:17 +0000)]
handle superinterfaces correctly for raw types
aclement [Wed, 20 Jul 2005 16:07:25 +0000 (16:07 +0000)]
generics: testcase for decp'ing parameterized type onto a generic type that implements a parameterized type using a type variable type parameter ;)
aclement [Wed, 20 Jul 2005 16:02:45 +0000 (16:02 +0000)]
commented out debug line...
acolyer [Wed, 20 Jul 2005 12:35:59 +0000 (12:35 +0000)]
ensure that when parameterizing a member, we handle any types that are themselves parameterized (eg. a return type List<N> should turn to List<String> when parameterized with String).
acolyer [Wed, 20 Jul 2005 11:11:29 +0000 (11:11 +0000)]
changes to construction and representation of parameterized types, and enforcement of the constraint that type mungers always operate on the generic type (never raw or parameterized)
aclement [Tue, 19 Jul 2005 07:08:32 +0000 (07:08 +0000)]
organized imports - wanted to trigger a build :)
avasseur [Mon, 18 Jul 2005 14:58:41 +0000 (14:58 +0000)]
ltw reweavable test for Mattew bug, fix some wrong getCoreType that makes some stderr output when closure are encountered on ltw / dump
avasseur [Mon, 18 Jul 2005 11:23:32 +0000 (11:23 +0000)]
fix LTW for JMX stuff, rename messageHolder, some doc on that + dump
aclement [Mon, 18 Jul 2005 09:45:21 +0000 (09:45 +0000)]
method name in weaver refactored.
aclement [Mon, 18 Jul 2005 09:36:00 +0000 (09:36 +0000)]
More bits of fix for pr98901 (annotation copying)
avasseur [Mon, 18 Jul 2005 09:26:20 +0000 (09:26 +0000)]
@DeclareAnnotation and guard to not use it in M3, doc update for M3
aclement [Mon, 18 Jul 2005 08:44:04 +0000 (08:44 +0000)]
First bits of fix for pr98901 (annotation copying)
aclement [Mon, 18 Jul 2005 08:34:11 +0000 (08:34 +0000)]
testcases for pr98901 (annotations copied to targets of decannotation). Not yet wired into the full 150 suite.
aclement [Mon, 18 Jul 2005 08:27:44 +0000 (08:27 +0000)]
Fix for pr103266: allow for nulls in the visitor (can happen if there are syntax errors)
acolyer [Fri, 15 Jul 2005 18:12:53 +0000 (18:12 +0000)]
moves forGenericType from UnresolvedType to ResolvedType. Return value still needs changing... one step at a time!
acolyer [Fri, 15 Jul 2005 18:11:34 +0000 (18:11 +0000)]
makes fromBinding in EclipseFactory an instance rather than static method. This paves the way for fromBinding to return a ResolvedType (which needs the world that the instance contains) rather than an UnresolvedType.
acolyer [Fri, 15 Jul 2005 16:09:02 +0000 (16:09 +0000)]
This is the beginning of a significant refactoring needed to cleanly support generics. We need to be much crisper about when we have an unresolved type (plain old TypeX) and when we have a resolved type. This refactoring renames TypeX to UnresolvedType and ResolvedTypeX to ResolvedType. In addition, it moves a number of methods that were defined on TypeX but which can only be answered by resolved types down onto ResolvedType. At the moment, ResolvedType still extends UnresolvedType - but a ResolvedType is *not* a UnresolvedType and this inheritance of convenience will be broken down the line so that ResolvedType does not extend UnresolvedType. Full signature information can only be known for ResolvedTypes.
acolyer [Thu, 14 Jul 2005 15:59:16 +0000 (15:59 +0000)]
just comments (honest ;) ) describing the format of signatures for TypeXs.
aclement [Wed, 13 Jul 2005 15:33:33 +0000 (15:33 +0000)]
organized imports
acolyer [Wed, 13 Jul 2005 13:04:31 +0000 (13:04 +0000)]
beginnings of support for matching against type variables
acolyer [Wed, 13 Jul 2005 13:04:03 +0000 (13:04 +0000)]
avoid losing type variable information when resolving bindings
acolyer [Wed, 13 Jul 2005 13:03:35 +0000 (13:03 +0000)]
now remembers its location for better error reporting
acolyer [Wed, 13 Jul 2005 13:02:54 +0000 (13:02 +0000)]
test case updates...
acolyer [Wed, 13 Jul 2005 13:02:35 +0000 (13:02 +0000)]
better error reporting when using type vars in inappropriate locations
acolyer [Wed, 13 Jul 2005 13:02:06 +0000 (13:02 +0000)]
added an additional message operation to tidy up weaver code
acolyer [Wed, 13 Jul 2005 13:01:37 +0000 (13:01 +0000)]
test for parameterized type usage
acolyer [Wed, 13 Jul 2005 13:00:59 +0000 (13:00 +0000)]
new message "a parameterized type pattern may not be used in a handler pointcut expression"
acolyer [Wed, 13 Jul 2005 13:00:08 +0000 (13:00 +0000)]
sometimes we get asked to resolve more than once, and that's ok...
acolyer [Wed, 13 Jul 2005 12:59:45 +0000 (12:59 +0000)]
more test cases covering generics in pointcut expressions
acolyer [Wed, 13 Jul 2005 10:13:38 +0000 (10:13 +0000)]
don't re-resolve already resolved types
acolyer [Wed, 13 Jul 2005 10:13:10 +0000 (10:13 +0000)]
teach a typex whether or not it represents a type variable
aclement [Wed, 13 Jul 2005 08:40:51 +0000 (08:40 +0000)]
organized imports
wisberg [Tue, 12 Jul 2005 21:00:30 +0000 (21:00 +0000)]
Ant 1.6 antlib.xml support, per bug 103493
acolyer [Tue, 12 Jul 2005 14:24:51 +0000 (14:24 +0000)]
handles resolution of bindings when the pointcut uses type variables
acolyer [Tue, 12 Jul 2005 14:24:13 +0000 (14:24 +0000)]
a pointcut now has a simple list of type variable names in scope rather than a TypeVariableListPattern
acolyer [Tue, 12 Jul 2005 14:23:35 +0000 (14:23 +0000)]
test that a signature with a generic declaring type does not match on the erasure of the parameter types
aclement [Tue, 12 Jul 2005 13:59:06 +0000 (13:59 +0000)]
organized imports
aclement [Tue, 12 Jul 2005 13:58:17 +0000 (13:58 +0000)]
generics: new error messages.
acolyer [Tue, 12 Jul 2005 13:19:22 +0000 (13:19 +0000)]
I needed to make some changes to resolveBindings, but the original method had degenerated into such a mess that I've refactored it before continuing. This commit changes no functionality, it just makes the code more readable!
acolyer [Tue, 12 Jul 2005 11:09:22 +0000 (11:09 +0000)]
parameterized members remember their erasure and this is used in member matching so that a signature pattern based on the erasure of a signature can match parameterized member declarations.
acolyer [Tue, 12 Jul 2005 11:08:09 +0000 (11:08 +0000)]
test that a the erasure of a signature matches a parameterized implementation of a method
wisberg [Mon, 11 Jul 2005 23:52:53 +0000 (23:52 +0000)]
tests for Huff's fix for 102746 - not hooked up b/c harness manages classpath.
wisberg [Mon, 11 Jul 2005 23:51:52 +0000 (23:51 +0000)]
Variant of Huff's fix for 102746 - untested hypothesis
avasseur [Mon, 11 Jul 2005 15:09:52 +0000 (15:09 +0000)]
99861 temp fix for deadlock when Sun' Unsafe.defineClass API is used (deadlock in JVMTI obviously, not tied to AJ, but bothers Ron)
avasseur [Mon, 11 Jul 2005 15:05:44 +0000 (15:05 +0000)]
fix 83935 where Jp / Pjp is an arg and bound in a formal binding as well as another (or 2+) used as implicit bindings in @AJ
acolyer [Mon, 11 Jul 2005 13:50:39 +0000 (13:50 +0000)]
add unresolved TypeVariableRefTypes which are created by EclipseFactory during conversion of bindings to TypeX, and resolved to TypeVariableReferenceTypes by the world.
acolyer [Mon, 11 Jul 2005 13:49:26 +0000 (13:49 +0000)]
a simple type pattern for a generic type should match any parameterization of that type
acolyer [Mon, 11 Jul 2005 13:48:57 +0000 (13:48 +0000)]
added one simple test for matching members of a parameterized type.
Took me 3 days to get it to pass!!!
acolyer [Mon, 11 Jul 2005 13:48:13 +0000 (13:48 +0000)]
changed EclipseFactory.makeResolvedMember from static to non-static member and fully resolved all types involved.
aclement [Mon, 11 Jul 2005 10:53:01 +0000 (10:53 +0000)]
generics: use raw type of iterator
aclement [Mon, 11 Jul 2005 10:52:46 +0000 (10:52 +0000)]
generics: modified to use built in types since otherwise world setup is painful.
aclement [Mon, 11 Jul 2005 10:52:22 +0000 (10:52 +0000)]
generics: temporarily use the raw type when matching until the type pattern logic is aware
aclement [Mon, 11 Jul 2005 10:51:48 +0000 (10:51 +0000)]
generics: polices the combinations of decp when using generics/raws/parameterized types
aclement [Mon, 11 Jul 2005 10:51:20 +0000 (10:51 +0000)]
generics: writes out correct class signature
aclement [Mon, 11 Jul 2005 10:51:05 +0000 (10:51 +0000)]
generics: improved type resolution, aware of raw/simple/parameterized/generic
aclement [Mon, 11 Jul 2005 10:50:35 +0000 (10:50 +0000)]
generics: dont assume bcelmethod, it cant be resolvedmember
aclement [Mon, 11 Jul 2005 10:50:15 +0000 (10:50 +0000)]
generics: signature unpacking.
aclement [Mon, 11 Jul 2005 10:48:54 +0000 (10:48 +0000)]
generics: dont assume its a bcelmethod. When working with generic methods they are merely 'resolvedmembers'
aclement [Mon, 11 Jul 2005 10:48:28 +0000 (10:48 +0000)]
generics: now promotes simple types to raw types if we discover they are generic during resolution.
aclement [Mon, 11 Jul 2005 10:47:42 +0000 (10:47 +0000)]
generics: copes with '+' '-' in the signature, less than ideal...
aclement [Mon, 11 Jul 2005 10:47:08 +0000 (10:47 +0000)]
generics: allow for type variables that arrive in static methods and don't come from the declaring type
aclement [Mon, 11 Jul 2005 10:46:30 +0000 (10:46 +0000)]
generics: few bug fixes to ensure typex's know if they represent generics/raws or simples.
aclement [Mon, 11 Jul 2005 10:46:02 +0000 (10:46 +0000)]
generics: lots of generic decp tests.
aclement [Mon, 11 Jul 2005 10:45:27 +0000 (10:45 +0000)]
generics: ask for the appropriate TypeX depending on the binding
acolyer [Mon, 11 Jul 2005 09:40:14 +0000 (09:40 +0000)]
but i only changed a comment...
acolyer [Mon, 11 Jul 2005 09:38:48 +0000 (09:38 +0000)]
fixes dumb array cast
acolyer [Mon, 11 Jul 2005 09:38:13 +0000 (09:38 +0000)]
ability to replace upper bound (used when fixing up circular dependencies)
acolyer [Mon, 11 Jul 2005 09:37:21 +0000 (09:37 +0000)]
test case for Enum conversions
acolyer [Mon, 11 Jul 2005 09:36:49 +0000 (09:36 +0000)]
answer correct superclass and superinterfaces taking into account generic signature
acolyer [Mon, 11 Jul 2005 09:35:36 +0000 (09:35 +0000)]
handle recursive type variables such as Enum
acolyer [Mon, 11 Jul 2005 09:34:44 +0000 (09:34 +0000)]
beginnings of proper type variable support in Eclipse type world
acolyer [Mon, 11 Jul 2005 09:32:20 +0000 (09:32 +0000)]
do the right thing with ":" in an ftp
aclement [Fri, 8 Jul 2005 16:31:44 +0000 (16:31 +0000)]
generics: includes fix for generic signatures