[
https://issues.apache.org/jira/browse/GROOVY-10352?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Eric Milles updated GROOVY-10352:
---------------------------------
Description:
Consider the following:
{code:groovy}
@AnnotationCollector(mode = AnnotationCollectorMode.PREFER_EXPLICIT_MERGED)
@interface Example {
}
{code}
{code:groovy}
@Example
class Sample {
}
{code}
When the compiler accesses the annotations of a pre-compiled Example through
the {{ClassNode}} Sample, the {{AnnotationCollector}} {{mode}} attribute is
null. This can be problematic for AST transformations which rely on the Groovy
model for the annotation metadata.
In my real-world example, I had a {{record}} type, which has the implicit
annotation {{RecordType}}. When I added an explicit {{TupleConstructor}}
annotation, the attributes were not merged, so I received duplicate processing
of {{TupleConstructor}} and therefore duplicate constructors.
was:
Consider the following:
{code:groovy}
@AnnotationCollector(mode = AnnotationCollectorMode.PREFER_EXPLICIT_MERGED)
@interface Example {
}
{code}
{code:groovy}
@Example
class Sample {
}
{code}
When the compiler accesses the annotations of Example through the {{ClassNode}}
Sample, the {{AnnotationCollector}} {{mode}} attribute is null. This can be
problematic for AST transformations which rely on the Groovy model for the
annotation metadata.
In my real-world example, I had a {{record}} type, which has the implicit
annotation {{RecordType}}. When I added an explicit {{TupleConstructor}}
annotation, the attributes were not merged, so I received duplicate processing
of {{TupleConstructor}} and therefore duplicate constructors.
> Java8 does not load enum values into annotation attributes
> ----------------------------------------------------------
>
> Key: GROOVY-10352
> URL: https://issues.apache.org/jira/browse/GROOVY-10352
> Project: Groovy
> Issue Type: Improvement
> Components: Compiler
> Affects Versions: 4.0.0-beta-2
> Reporter: Eric Milles
> Assignee: Eric Milles
> Priority: Major
>
> Consider the following:
> {code:groovy}
> @AnnotationCollector(mode = AnnotationCollectorMode.PREFER_EXPLICIT_MERGED)
> @interface Example {
> }
> {code}
> {code:groovy}
> @Example
> class Sample {
> }
> {code}
> When the compiler accesses the annotations of a pre-compiled Example through
> the {{ClassNode}} Sample, the {{AnnotationCollector}} {{mode}} attribute is
> null. This can be problematic for AST transformations which rely on the
> Groovy model for the annotation metadata.
> In my real-world example, I had a {{record}} type, which has the implicit
> annotation {{RecordType}}. When I added an explicit {{TupleConstructor}}
> annotation, the attributes were not merged, so I received duplicate
> processing of {{TupleConstructor}} and therefore duplicate constructors.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
