[GROOVY-7940] Inconsistent handling of undocumented primitive data types at the @Lazy annotation under use of extends

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.4.5
• Fix Version/s: 2.4.8
• Component/s: groovy-jdk
• Labels: data extends lazy primitives super types
• Environment: Any
• Assignee: John Wagenleitner
• Reporter: Maik Igloffstein
• Created: 14/Sep/16 12:45
• Updated: 01/Feb/17 23:18
• Resolved: 18/Sep/16 05:33

Description

Documentation: http://docs.groovy-lang.org/latest/html/gapi/groovy/lang/Lazy.html

The documentations describes the usage of @Lazy like:
@Lazy T x

I'm aware of that T means no primitive data types, but groovy allows it and generates the code - so why not.

class Super {
    boolean aBoolean = true
    int aInt = 5
}

class Testing extends Super {

    @Lazy
    boolean aBoolean = { false }()
    
    @Lazy
    int aInt = { 5 * 5 }()

}

println new Testing().isaBoolean()
println new Testing().getaBoolean()
println new Testing().getaInt()
println new Testing().aBoolean
println new Testing().aInt

Result:

true
false
25
true
25

The behavior is inconsistent.
boolean: results are from the super class
int: results are from the child class

Please fix it or throw a 'RuntimeException' when primitive data types are used.
Update:
Please let @Lazy implement _is_Boolean() instead of _get_Boolean() when using primitive data type boolean and document it or just throw a 'RuntimeException'.

Issue Links

• links to: GitHub Pull Request #420