Catch warning from unsupported hierarchy entries

[orangejulius]

During an import I noticed an error like the following:

error TypeError: Cannot read properties of undefined (reading 'name')
  at /code/import/source/whosonfirst/map/hierarchies.js:26:62
  at Array.forEach (<anonymous>)
  at Object.mapper [as hierarchies] (/code/import/source/whosonfirst/map/hierarchies.js:24:15)
  at mapper (/code/import/source/whosonfirst/map/place.js:39:7)
  at DestroyableTransform._transform (/code/import/mapperStream.js:7:21)
  at Transform._read (/code/node_modules/through2/node_modules/readable-stream/lib/_stream_transform.js:166:10)
  at Transform._write (/code/node_modules/through2/node_modules/readable-stream/lib/_stream_transform.js:155:83)
  at doWrite (/code/node_modules/through2/node_modules/readable-stream/lib/_stream_writable.js:390:139)
  at writeOrBuffer (/code/node_modules/through2/node_modules/readable-stream/lib/_stream_writable.js:381:5)
  at Writable.write (/code/node_modules/through2/node_modules/readable-stream/lib/_stream_writable.js:302:11)

After some investigation these errors don't appar to be hurting anything, they happen when a record in a layer we don't currently support is imported. In this case it was the postalregion layer.

We could consider supporting that layer in the future, which I think is also easy, but in the meantime we should at least silence these errors.

[missinglink]

Nice catch.

I like the idea of not necessarily silencing the warnings but at least catching them and handling them more gracefully.

One potential issue with this is that I just merged this PR which introduced the concept of the ontology.type not exactly matching the placetype (or equivalent).

As such this change could potentially become brittle to failure if that ontology mapping concept became more widespread, and particularly if we decided to adopt an ontology not directly mirroring WOF.

That said, I'm fine with this for now if we don't want to over-think it.

Alternatively I'd also be fine with checking $pt and bailing with a warning.

[missinglink]

Worth noting that, in theory, the WOF spec should contain all valid entries (such as postalregion) however that doesn't seem to be the case. I've also flagged others in the past, I think it's just not up-to-date.

[missinglink]

I ran the update script in import/source/whosonfirst/config and it seems to have updated the config to include the latest additions to the WOF placetypes.

jq -r 'to_entries[] | "\(.key): \(.value.name)"' import/source/whosonfirst/config/placetypes-spec-latest.json | grep post
470996387: postalcode
1897483759: postalregion

[missinglink]

On reflection I think this looks good, we can handle the schema update in a separate PR