Bug or expected behaviour?

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Bug or expected behaviour?

Keith Suderman

I have a parent class that defines a static final String and a child class that references the String in it's constructor.  Essentially:

class Parent {
        static final String DEFAULT = 'default'
class Child extends Parent {
        String value
        Child() { this(DEFAULT) }  // Bad type on operand stack
        Child(String value) { this.value = value }
println new Child().value
However, when I try to run the code I get a "java.lang.VerifyError: Bad type on operand stack" (see below).  From my understanding this exception is caused when a constructor tries to use a variable/field that has not been initialized yet; i.e. the `this(DEFAULT)` in the above example.  However, if I qualify the DEFAULT with the class name (i.e. `this(Parent.DEFAULT)`) the code works as expected.

Similar code in Java works without DEFAULT being qualified with the superclass name.  So I am wondering; is this a bug or a design decision that was made somewhere/sometime?  

I am using Groovy Version: 2.5.2 JVM: 1.8.0_40 Vendor: Oracle Corporation OS: Mac OS X


Caught: java.lang.VerifyError: Bad type on operand stack
Exception Details:
    Child.<init>()V @10: invokeinterface
    Type uninitializedThis (current frame, stack[2]) is not assignable to 'java/lang/Object'
  Current Frame:
    bci: @10
    flags: { flagThisUninit }
    locals: { uninitializedThis, '[Lorg/codehaus/groovy/runtime/callsite/CallSite;' }
    stack: { uninitializedThis, 'org/codehaus/groovy/runtime/callsite/CallSite', uninitializedThis }
    0x0000000: b800 114c 2a2b 1212 322a b900 1802 00b8
    0x0000010: 001e c000 20b7 0023 b1