What's the best way to define both geo shape and geo point dynamic mapping

Elastic Stack Elasticsearch
1

Using dynamic templates, I have a "locations" mapping for "location" field with mapping type "geo_point". Some visualizations need this field to be a "geo_shape", however. I would like to extend the mapping similar to how we'd add "fields": { "raw" ... } to support both but unsure how best to define a single mapping that can include both "geo_shape" and "geo_point".

Guess is something like:

"dynamic_templates": [
  {
     "locations": {
       "match": "location",
       "mapping": {
           ???
       }
     }
  }
]

I'm not sure if you can just set mapping type to "geo_point" and have fields with both types or better to define two completely separate mappings and matches with different field naming.

I want both point and shape:

"type": "geo_point"

and

"type": "geo_shape",
"tree": "quadtree",
"precision": "50m"

Would something like this work or better yet, is there a way to use copy_to to copy the lat/lon to both on single insert?

  "dynamic_templates": [
    {
      "locations": {
        "match": "location",
        "mapping": {
          "type": "geo_shape",
          "tree": "quadtree",
          "precision": "50m",
          "fields": {
            "point": {
              "type": "geo_point"
            }
          }
        }
      }
    }
  ]

Appreciate tips on best practice here.

After some trial/error I learned the following so wonder if conclusion is you have to add different fields and types and populate them all redundantly on insert.

  1. copy_to will not copy to a multi-field type (geo_point)
  2. mapping will not allow nested fields

So far I got this to work (all separate fields) but there has to be a more elegant way to store these without the redundant entry. https://github.com/mikesparr/elasticsearch-geo-template

2

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.