Hi Folks,
I am taking a look at jchord (repository head) to determine what kind of
analyses are possible with it. I started with writing a small callgraph to
dot converter to see the differences between the different callgraph
generation settings (ci, cs, co, cc and so on).
I was able to implemented it very quickly (great compliment for a very good
framework ... even it still lacks documentation). But I had some problems
with the hashCode implementation of several classes (Ctxt and the tuple
objects).
For each method in my demo application I collect a Set<Ctxt> from which the
method is called. Since the resulting dot graph was curious I took a closer
look and I observed that logical different context produce the same
hashcode (In my case all context from calls to java.lang.Object.<init>
produce a hashCode of 1). Is this intended?
Bernhard
Original issue reported on code.google.com by berge...@googlemail.com on 31 May 2010 at 6:53
Original issue reported on code.google.com by
berge...@googlemail.comon 31 May 2010 at 6:53