2010/11/4 Marko Rodriguez <okramma...@gmail.com>
> Hi Mattias,
>
> Actually: one more question. Is the uniqueness of an index still dependent
> on equality of Map<String,String> configuration?
>
> In other words can I do this?
>
> Index i = forNodes("myIndex", someMap)
> Index j = forNodes("myIndex")
>
> assertTrue(i,j)? (this statement is not to be taken literally, as I don't
> want to test equality of indices in this manner, but I hope you get the
> point I'm trying to effect).
>
i and j would be the same, yes. It's the entityType (node/relationship) +
index name which makes 'em unique, not the configuration).
>
> Thanks,
> Marko.
>
> On Nov 3, 2010, at 4:17 PM, Marko Rodriguez wrote:
>
> > Perfect on both. Thanks!
> >
> > Marko.
> >
> > On Nov 3, 2010, at 4:15 PM, Mattias Persson wrote:
> >
> >> 2010/11/3 Marko Rodriguez <okramma...@gmail.com>
> >>
> >>> Hello,
> >>>
> >>> Here is what I would like to request of the new index model.
> >>>
> >>> 1. Ability to getAllIndices().
> >>>
> >>
> >> There's the:
> >>
> >> graphDb.index().nodeIndexNames();
> >> graphDb.index().relationshipIndexNames();
> >>
> >> methods, you're thinking of something else?
> >>
> >>
> >>> 2. Ability to add metadata to indices that is saved to disk (for
> >>> Blueprints)
> >>> - need to note if an index is an automatic index or manual index.
> >>> - if its an automatic index, need to be able to specify which keys
> >>> are being indexed.
> >>>
> >>
> >> There's a shell app which exposes ways of modifying index configuration
> >> parameters... of course it's not of any use to Gremlin/Tinkerpop, but
> I'm
> >> just saying it's possible to do that, and maybe such methods will find
> its
> >> way up to the API (although it's very advanced stuff to tamper with
> index
> >> configuration in the middle of a lifetime of an index, since it could
> affect
> >> index storage format a.s.o.).
> >>
> >> But if there would be a setConfigurationParameter method or something
> like:
> >>
> >> Index<Node> myIndex = ...;
> >> myIndex.setConfigurationParameter( "keys", "name,title,whatever" );
> >>
> >> it would do what you're asking for, right? Because this exists today,
> just
> >> not exposed in a method like that.
> >>
> >>
> >>>
> >>> 2 is gnarly.. Is there any solution anyone can think of (without
> >>> introducing vertices that, for example, maintain this information)?
> >>>
> >>> Thanks,
> >>> Marko.
> >>>
> >>> http://markorodriguez.com
> >>> _______________________________________________
> >>> 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
>
--
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
