Re: gcj/257 (workaround)

[Alexandre Petit-Bianco <apbianco at cygnus dot com>]

The following reply was made to PR gcj/257; it has been noted by GNATS.

From: Alexandre Petit-Bianco <apbianco@cygnus.com>
To: java-gnats@sourceware.cygnus.com
Cc:  
Subject: Re: gcj/257 (workaround)
Date: Fri, 21 Jul 2000 22:44:40 -0700 (PDT)

 Mark Wielaard writes:
 
 >  As a workaround you can cast to (B$C) which is accepted by gcj.
 >  (But not by jikes and it seems wrong that gcj accepts it.)
 
 It is, because `$' isn't a qualified expression separator. Jikes is
 right to reject it and most likely accepts something like:
 
   class A$B {}
   class C { A$B x = null;}
 
 Compilers of some versions of the JDK accept it interesting ways:
 jdk117 accepts the following:
 
   class a$b {}
   class z {
     Z$C x = null;
     a$b y = null;
   }
 
   class Z {class C {}}
 
 I.e. will accept a class called a$b and will take the liberty to
 interpret Z$C as a qualified expression to point at inner class C.
 
 In the light of these facts, I'm going to amend my patch to 257.
 
 ./A