|
From: Peter G. <gilesp@u.washington.edu> - 2008-08-18 17:04:19
|
Excellent! I'm very pleased to be able to contribute. The BitSet suggestion is a good one, an obvious winner. Wish I had thought of it ;) Cheers, Peter On Sun, 17 Aug 2008 01:22:45 -0700, "John V. Sichi" <js...@gm...> said: > John V. Sichi wrote: > > Thanks Peter, this is a really useful contribution. I'll work on > > including it in the next release; in the meantime, I'm cc'ing it > > to jgrapht-users in case anyone wants to make use of the code > > before then. > > I have checked this into Subversion under new package > org.jgrapht.experimental.dag, which is where it will go for the 0.7.4 > release. After that, it can move out of experimental once we've > determined the best place for the various interfaces (we don't > currently have a pattern of a constrained graph within JGraphT, so we > might want to add infrastructure for the constraint concept > generically first). > > Also, I'm thinking that a java.util.BitSet could provide a good > implementation for the Visited interface (1/8 the memory of a > boolean array). > > JVS |
