[BUG] Decompiler fails on interpreting case inside block

[marcus at ira.uka.de]

Ned, could you check if this fixes the problem?

Subject:		[FIX] decompFixes-ls
Author:		Lex Spoon
Date Posted: 30 January 2004
Archive ID: 19050
Comments:	
from preamble:

"Change Set:		decompFixes-ls
Date:			30 January 2004
Author:			Lex Spoon

Fixes the decompiler bug posted by Ned Konz on 11 February 2003.  The
issue is that if the last case in a case statement returns from the
method, the decompiler gets confused trying to parse the otherwise:
branch.  This changeset has the decompiler track all non-return exits
from the case statement, so that it has enough information to do a right
thing.