Change GADM search API about gbif-api HOT 4 CLOSED

Search params

An additional search parameter gadmGid ...

I too had imagined having a catch all GID param like that. The GIDs are unique as far as I know, it seems redundant to specify the level in search. It also has the benefit that I can search for say Denmark and Malmø as they are OR'ed (Similar to above example I believe).

Facets
As for facets then I'm in favour of both a facet=gadmGid (similar to facet=taxonKey) and a facet=gadmLevel2Gid (similar to facet=phylumKey)

Search by name
Keeping search by name is useful if we want to do suggests driven of ES for scoped portals.
It is then possible to do a search by name and facet by gadmGid and use those facets as suggestions. I isn't a high priority for now - but we loose that possibility if we do not include names in the index.

country vs gadm
Perhaps I do not appreciate the complexity, but I do not see any reason to do anything special with ISO country vs GADM country. They are different definitions and should live separate lives I would think.

from gbif-api.

Thanks.

Keeping search by name is useful if we want to do suggests driven of ES for scoped portals.
It is then possible to do a search by name and facet by gadmGid and use those facets as suggestions. I isn't a high priority for now - but we loose that possibility if we do not include names in the index.

Is that needed? We don't have a facet way to return kingdom names. We will provide a lookup service from GID→name, and the result can be cached indefinitely.

There are duplicate area names in neighbouring countries, e.g. there's a "Berg" municipality in both Norway and Sweden.

from gbif-api.

No I do not believe names are needed in the occurrence index - not with @fmendezh suggestion for an autocomplete where suggestions can be scoped geographically (only give me suggestions for locations within this GID). If so it would only be for the benefit of unstructured search ?q=Ringsted

from gbif-api.

This is ready in UAT.

from gbif-api.