Field initializer anonymous inner class fouling up type inferencing / semantic highlighting

[mauromol]

Follow-up to #378.
Consider the following Groovy class:

package test8
class A {
	protected def m() { }
	
	Object p = new Object() {
		String toString() {
			m()
		}
	}
	
	void init() {
		def whatever = new Object() {
		  def something() {
			m()
		  }
		}
	  }
  }

m() call from the definition of whatever local variable is not underlined, but it's not actually recognized, as you can see with F2 or F3 on it, or by trying to rename m() to something else.

Please note that if you comment out the p field definition, then the second call to m() becomes recognized.

[eric-milles]

ready to test

[mauromol]

Hi Eric, I tested this with 2.9.2.xx-201711130408-e46 and it seems to be fixed, thank you! 👍