Closed missinglink closed 7 years ago
@dphiffer This edit came from Boundary Issues and produced a string property value (should be an int). Are we missing an entry in a schema config somewhere that says this property should only have ints?
Updated property_types file in https://github.com/whosonfirst/whosonfirst-tests/commit/6bb1f792a48dc7c7d3f049bdb0883988dc94afbf to include wof:population
and wof:population_rank
(both as "int" types).
@dphiffer after @stepps00 change in https://github.com/whosonfirst/whosonfirst-tests/commit/6bb1f792a48dc7c7d3f049bdb0883988dc94afbf do you need to update anything in Boundary Issues to pickup that schema change?
I just added wof:population
and wof:population_rank
to the schema, which I think should help. (see: https://github.com/whosonfirst/whosonfirst-www-boundaryissues/commit/f5e885eb384986e4d05bbaf4654c787a57514b53)
There is more to this issue, unfortunately, but maybe not especially relevant here. Basically when the edit page loads it includes clues which about datatype is expected, embedded in the HTML. So now when we save the record with an existing wof:population
property it should take effect (see: https://github.com/whosonfirst-data/whosonfirst-data/commit/4aaeb12b8457d4ddc15b01267c98d18eb4f5c977).
Unfortunately the schema isn't consulted when a new property gets added to a record, so I still haven't totally addressed this in BI.
Closing this PR - the property json was updated in a prior commit to include these changes.
https://github.com/whosonfirst/whosonfirst-tests/blob/master/property_types.json#L47-L48
hey, I noticed the
wof:population
properties are being encoded as strings when they should probably be integers.