Fantastic. Very nice.
I guess this was also inspired by Marco's Gremlin and pipes?
(so the use case is now well understood?)
On Tue, Mar 15, 2011 at 11:33 AM, Mattias Persson <matt...@neotechnology.com
> wrote:
> Already though of that :) behold:
>
> PipeBuilder.fromNode(startNode).into(otherNode(A))
> .into(traverse(myTraversalDescription))
> .into(traverse(myOtherTraversalDescription))
> .into(otherNode(B));
>
> Or whatever. You can even use other "from", f.ex: fromNode, fromNodes,
> fromPath, fromPaths a.s.o.
>
> Something like that?
>
> 2011/3/15 Craig Taverner <cr...@amanzi.com>
>
> > I like the pipes idea. What I would like to see is nested traversers. The
> > pipe example below seems to imply single hops at each step, but it would
> be
> > nicer to allow each step to traverse until it reached a certain criteria,
> > at
> > which point a different traversal would take over.
> >
> > In the old and current API's it seems to do this you need to create a
> > traversal, iterate over it, and create a new traversal inside the loop.
> >
> > We created a Ruby DSL for nested traversals a year or so ago that looks a
> > bit like:
> >
> >
> > chart 'Distribution analysis' do
> > self.domain_axis='categories'
> > self.range_axis='values'
> > select 'First dataset',:categories=>'name',:values=>'value' do
> > from {
> > from {
> > traverse(:outgoing,:CHILD,1)
> > where {type=='gis' and name=='network.csv'}
> > }
> > traverse(:outgoing,:AGGREGATION,1)
> > where {name=='azimuth' and get_property(:select)=='max' and
> > distribute=='auto'}
> > }
> > traverse(:outgoing,:CHILD,:all)
> > end
> > end
> >
> > This is quite a complex example, but the key points are the from method
> > which defines where to start a traversal, and the traverse method which
> > defines the traversal itself, with the where method which is like the old
> > ReturnableEvaluator.
> >
> > Will the new pipes provide something like this?
> >
> > On Tue, Mar 15, 2011 at 9:19 AM, Massimo Lusetti <mluse...@gmail.com>
> > wrote:
> >
> > > On Tue, Mar 15, 2011 at 9:11 AM, Mattias Persson
> > > <matt...@neotechnology.com> wrote:
> > >
> > > > I'm positive that some nice API will enter the kernel at some point,
> > > f.ex.
> > > > I'm experimenting with an API like this:
> > > >
> > > > for(Node node :
> > > >
> > >
> >
> PipeBuilder.fromNode(startNode).into(otherNode(A)).into(otherNode(B)).nodes())
> > > > {
> > > > // node will be (3) from the example above
> > > > }
> > > >
> > > >
> > > > I hope I didn't confuse you with all this :)
> > >
> > > Nope, the opposite. Thanks for the clarification and that kind of API
> > > would be a killer feature IMHO.
> > >
> > > It will be even more pleasant to work with neo4j...
> > >
> > > Cheers
> > > --
> > > Massimo
> > > http://meridio.blogspot.com
> > > _______________________________________________
> > > Neo4j mailing list
> > > User@lists.neo4j.org
> > > https://lists.neo4j.org/mailman/listinfo/user
> > >
> > _______________________________________________
> > Neo4j mailing list
> > User@lists.neo4j.org
> > https://lists.neo4j.org/mailman/listinfo/user
> >
>
>
>
> --
> Mattias Persson, [matt...@neotechnology.com]
> Hacker, Neo Technology
> www.neotechnology.com
> _______________________________________________
> Neo4j mailing list
> User@lists.neo4j.org
> https://lists.neo4j.org/mailman/listinfo/user
>
_______________________________________________
Neo4j mailing list
User@lists.neo4j.org
https://lists.neo4j.org/mailman/listinfo/user
