Hi Aseem,
This is a use case that auto indexing may not be the best fit. The
purpose of auto indexing is taking care of the common use case in the
simplest possible manner, which is a property added is a property
indexed. Indexing a property with a different value that is present on
the primitive (that is what I understood from your description,
correct if wrong) is a case that is best suited for custom indexing
operations. Additionally, it is a desirable quality to have an index
which follows the contents of the actual database with guaranteed
consistency. I would be interested to know if there are others that do
similar transformations and in what manner - maybe there is room for
providing a customizing API.
cheers,
CG
On Sat, Jul 16, 2011 at 9:48 AM, Aseem Kishore <aseem.kish...@gmail.com> wrote:
> We transform some of our properties currently before (manually) indexing
> them. Simple examples are lowercase ("fulltext" indexes supports a
> "lowercase" config that defaults to true, but "exact" ones don't, I was
> disappointed to find), but more complex ones are e.g. stripping special
> characters.
>
> It would be great if the auto-indexer could support this notion of
> transforming property values (and queries) through user-defined function,
> just like traversal accepts a user-defined filter function.
>
> (And, fwiw, we're users of the REST API, so ideally this functionality would
> be exposed over REST too.)
>
> Cheers,
> Aseem
> _______________________________________________
> 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
