I think the key point of Peters request is to separate the 'builder' from the 'traverser'. Mattias argument appears to state that if the builder and traverser are the same class (series of immutable instances of the same class), you have more flexibility in refactoring, because you don't have to remember to convert to the traverser as the last step (Peter called build() as the last step of a builder creating a 'really immutable' traverser).
The example of String.replace is similar to Mattias description, one immutable object creating another instance of the same class. However, I personally think Peter has a point. Keeping the builder and traverser as two separate objects might force the coder to make a clear decision about when the traverser should no longer be modified. Once the build() method is called, we are done. Before that point the developer still has complete flexibility to modify the traversal rules multiple times in as many steps and places in the code they want, but once they call build(), the returned traverser cannot be modified at all. Despite my argument above favouring Peters approach, I actually do not have a strong opinion on this, and think both ways are good. I confess that had I coded this myself there is a good way I would have done it with one class like Mattias describes. (and I just realized another thing, we normally access traversers in for(Iterable) loops that call iterator(), which itself returns a 'really immutable' object, so perhaps adding another in the chain is overkill :-) On Tue, Jul 27, 2010 at 9:00 PM, Paul A. Jackson <[email protected]>wrote: > I may have missed your point. But, FWIW, this model reflects what I would > expect from an immutable object. > For example: > String s = "Test"; > s.replace('T', 't'); // s still contains "Test" > > BigInteger and Date are the same way. > > -Paul > > > > -----Original Message----- > From: [email protected] [mailto:[email protected]] > On Behalf Of Mattias Persson > Sent: Tuesday, July 27, 2010 2:41 PM > To: Neo4j user discussions > Subject: Re: [Neo4j] TraversalDescription building hickup > > 2010/7/27 Peter Neubauer <[email protected]> > > > Hi all, > > I just stumbled over the immutable TraversalDescription API > > (http://components.neo4j.org/neo4j-kernel/apidocs/index.html), which > > will not modify the object if you do > > > > TraversalDescription td = new TraversalDescriptionImpl(); > > td.depthFirst(); > > > > Instead, one needs to reassign td, like > > > > TraversalDescription td = new TraversalDescriptionImpl(); > > td = td.depthFirst(); > > > > However, > > TraversalDescription td = new TraversalDescriptionImpl().depthFirst(); > > > > will give you the expected td. > > > > IMHO this is unexpected behaviour and hard to get if you just follow > > the common fluent API and presume a Builder-pattern. Especially since > > no errors are thrown and you just end up with strange results and > > unreachable code i e.g. a custom PruneEvaluator etc. True, the API > > says it is immutable, but still I think this is hard. > > > > WDYT? > > > > Should we think of changing this to a proper builder.modify().modify > > etc and finally builder.build() wich gives you the final, immutable > > instance of TraversalDescription and is clearly understandable by > > clients? > > > > I still think the current approach is more useful (although it'd be nice > with more input on this). One reason I think it's better is that you can > half-bake descriptions as private static final or similar and then complete > the descriptions in several different places in your code. You can even > pass > in descriptions in methods and what not, without any risc of them being > modified. I think javadoc should better explain this and it should be > expected that developers read javadoc, right? > > > > > > Cheers, > > > > /peter neubauer > > > > COO and Sales, Neo Technology > > > > GTalk: neubauer.peter > > Skype peter.neubauer > > Phone +46 704 106975 > > LinkedIn http://www.linkedin.com/in/neubauer > > Twitter http://twitter.com/peterneubauer > > > > http://www.neo4j.org - Your high performance graph > database. > > http://www.thoughtmade.com - Scandinavia's coolest Bring-a-Thing party. > > _______________________________________________ > > Neo4j mailing list > > [email protected] > > https://lists.neo4j.org/mailman/listinfo/user > > > > > > -- > Mattias Persson, [[email protected]] > Hacker, Neo Technology > www.neotechnology.com > _______________________________________________ > Neo4j mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > > _______________________________________________ > Neo4j mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > _______________________________________________ Neo4j mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
