On 02.11.2010 14:52, Craig Taverner wrote:
>> Adding neo4j spatial to the picture should enable the CMS handling geo
>> data objects without adding to much complexity (I hope).
> OK. I think I'm beginning to get an idea of what you want. You have a CMS,
> and some of the data is geolocated, in the sense that you know what
> countries it is associated with, and perhaps you even have relationships to
> 'country' nodes (or other nodes) representing locations. Now you want to use
> Neo4j Spatial to perform simple spatial queries on this.
Your analysis is correct and brilliant. :-)

In my case, the CMS payload data itself (like HTML, Images, Videos, CSS, 
JS, Blogs etc.) is stored in neo4j nodes. But logging data, too, of course.

>   I could imagine something like:
>
> 'give me all countries in the specified region (bounding box) and include
> the number of pages as an attribute of the result set'bute?
Right now I'm using the SearchIntersectWindow class to determine which 
countries are needed to display within a certain global region. The 
countries data is loaded as a set of GeoJSON files via Javascript (using 
Polymaps), so neo4j spatial does only provide the country names.

> I think you should either use the original shapefiles and the
> existing ShapefileImporter or move to OSM data models for connected
> topologies.
>
> A GeoJSON importer is a nice idea, but not worth the effort if you already
> have shapefiles.
Indeed! I ended up with using the shapefile importer, using neo4j 
spatial to serve geodata nodes which are then connected on-the-fly with 
geo-tagged CMS data.

Your comments and suggestions helped me a lot, so I just wanted to say 
thanks again.

Axel



_______________________________________________
Neo4j mailing list
User@lists.neo4j.org
https://lists.neo4j.org/mailman/listinfo/user

Reply via email to