summaryrefslogtreecommitdiffstats
path: root/bcel-builder/src
Commit message (Collapse)AuthorAgeFilesLines
* add support and tests for isAnonymous in JavaClassacolyer2005-11-041-1/+19
|
* Fix for generic signature parsing - see pr112105aclement2005-10-141-0/+1
|
* pr112514: better diagnostics when class format exception occurs.aclement2005-10-141-2/+11
|
* final bcel fix for 101047acolyer2005-09-281-1/+4
|
* restoring original behaviour in LocalVariableGen - not happy with my change ↵acolyer2005-09-281-8/+2
| | | | from last night....
* fix for pr101407 - weaver produces wrong local variable table bytecode.acolyer2005-09-271-2/+9
|
* fix for pr104957: NPE whilst compiling GIJacolyer2005-09-211-35/+38
|
* tests and implementation for 108120 - runtime pointcut parsing and matching.acolyer2005-09-212-2/+35
|
* allow for references to descriptors when class_info structures refer to ↵aclement2005-09-191-2/+4
| | | | arrays. (pr109728)
* test and fix for pr109486: bcel exception when expecting compiler error for ↵aclement2005-09-141-4/+5
| | | | final abstract class.
* tests and fix for pr108062, failing to correctly tokenize (ZZLjava/...)acolyer2005-09-021-2/+4
|
* Fix for pr104720: busted tableswitch on around advice weaving.aclement2005-08-252-3/+14
|
* fix for pr107784, failure to parse an array of a primitive type when used as ↵acolyer2005-08-241-1/+7
| | | | a type variable
* better error messages when failing to unpack a generic signatureacolyer2005-08-241-4/+8
|
* for one little change, this sure was a hell of a lot of work to track down. ↵acolyer2005-08-121-2/+2
| | | | Fixes bug pr106130 where methods using iinc instructions with > 255 local vars were failing to generate wide versions of iinc instruction correctly.
* another case of ArrayStoreException - a bug that Java generics would ↵acolyer2005-08-121-1/+1
| | | | actually have prevented.
* fixed bug whereby we were adding a String into an array of Signature objectsacolyer2005-08-111-1/+1
|
* organized importsaclement2005-07-121-3/+1
|
* do the right thing with ":" in an ftpacolyer2005-07-111-1/+1
|
* fix in GenericSigParser when tokenising primitivesacolyer2005-07-081-0/+22
|
* generics: can determine if java class is generic.aclement2005-07-081-3/+10
|
* lazy loading and parsing of generic signatureacolyer2005-07-081-3/+21
|
* signature parsing for methods and fields now complete alsoacolyer2005-07-012-4/+7
|
* updates to Signature parsing for genericsacolyer2005-07-012-329/+520
|
* -- here be dragons --acolyer2005-06-261-2/+417
| | | | | | this is an early check-in for handover to Andy. I've added code to parse the generic signature attributes, but only completed parsing for class signature so far, method and field to do although these will reuse many of the same building blocks. Needs test cases and probably some more rework. Would not normally check-in at this stage. -- end of dragons --
* added getGenericSignature methodacolyer2005-06-261-1/+27
|
* renamed "getRealSignatureFromAttribute" to "getGenericSignature" (which is ↵acolyer2005-06-261-4/+4
| | | | what it is).
* Upgraded BCEL annotation support - can now get an immutable form of ↵aclement2005-06-0812-10/+88
| | | | annotation (type Annotation) from the mutable form (type AnnotationGen). Required when copying annotations between class files.
* GenericsWork: Angly aware signature processing.aclement2005-06-011-4/+26
|
* GenericsWork: Lots of changes in here, tidying stuff up, moving it around. ↵aclement2005-06-011-562/+577
| | | | Have made some of the signature processing generics aware (still work in progress) - also started adding new parsing code for the more complex signatures we are likely to encounter, this is tested but not currently used anywhere. See GenericSignatureParsingTest
* GenericsWork: Members now recognize that they can have a signature attribute ↵aclement2005-06-011-1/+35
| | | | attached that records the signature pre-erasure. It is accessible through 'getDeclaredSignature()' and is resolved lazily.
* Changed runtime invis param attrs to always be invis as discovered by Alex. ↵aclement2005-04-222-2/+2
| | | | Was impacting nothing as far as I could tell - but needed fixing!
* Changed runtime invis attrs to always be invis as discovered by Alex. Was ↵aclement2005-04-221-1/+1
| | | | impacting nothing as far as I could tell - but needed fixing!
* Declare annotation: infrastructure upgrade. Mainly to ensure we copy ↵aclement2005-03-1012-49/+116
| | | | entries between constant pools correctly - required when annotations move from one class to another (as is typical with declare annotation)
* For Martin Lippert - removed static state from class.aclement2005-03-102-117/+165
|
* More accessors for the different types of value.aclement2005-02-261-0/+50
|
* Removing last 'use the static type for accessing static field' warnings from ↵aclement2005-02-024-11/+11
| | | | the AJ source...
* Bit of missing support...aclement2005-01-312-2/+8
|
* @AJ syntax (pr83626): changes to bcel from Alexaclement2005-01-261-2/+5
|
* This fix for 'c' values in annotations has been verified by Pepe. ↵aclement2005-01-242-1/+11
| | | | Committing source changes and testcase.
* Allows for MarkerAnnotations - doh! (i.e. ones with no values) and unpacks ↵aclement2004-12-092-3/+10
| | | | invis annotations on classgens correctly.
* Fix some stuff that Wes pointed out: addEmptyConstructor() in ClassGen has ↵aclement2004-11-2241-187/+266
| | | | always been broken and the src.zip for BCEL that was generated included test imports.
* BCEL Java5 Supportaclement2004-11-193-0/+277
|
* BCEL Java5 Supportaclement2004-11-19342-514/+1393
|
* BCEL Java5 Supportaclement2004-11-1922-0/+1805
|
* Some enhancements discovered whilst working on bugzilla bug 78954: Compiler ↵aclement2004-11-185-22/+32
| | | | cannot cope with 4000 jars on the classpath. These enhancements cause things to be done 'lazily' in BCEL.
* Imports organized.aclement2004-11-1826-111/+153
|
* Adding BCEL src to bcel-builder project. We change/patch BCEL more ↵aclement2004-11-18342-0/+56495
| | | | | | frequently than we import new versions of BCEL into AspectJ - so it makes sense to make that patching process easier than the importing process. Also the Java5 support would result in an unwieldy patch.txt file. The ANT targets in build.xml will still enable transformation of the packaging from org.aspectj.apache to org.apache for diffing with a new BCEL version, should one ever appear.
* Removing this rogue directory that didn't seem to work. i.e. it didnt fix ↵aclement2004-01-301-0/+0
| | | | the build process.
* Following yesterdays discussion, I'm temporarily adding a folder containing ↵aclement2004-01-301-0/+0
one file to see if it makes the build happy (we had to do this for adrians new module). If it fails then I'll delete it and wait for George to return who knows the right cruise control incantations to locate the problem.