Two small fixes for things I ran into while testing the analysis changes.
The OSM extract download script wasn't quite right for US states. I hadn't noticed because it was using a cached version of the extract from my S3 bucket. Once I cleared that out, it started producing an error. The issue is that, whereas the country name in the URL for every other country (as far as I can tell) is the full name (in lower case, with hyphens for spaces), the country name for the US is just "us". This adds a condition in download_osm_extract.py to adjust for that.
A couple times I accidentally changed the new "Max Trip Distance" then cleared the field again to let it use the default, but then ended up with a crash because it got a null value rather than the default. I added a check to the front-end to send it as undefined (which is what it does if the field is never set) rather than null (which is what it ends up as if you set then clear it).
Notes
I'm putting this up as a PR for visibility, but these were in place for all the testing I just did, and I don't think we need another round of testing for them, so I'm just going to merge this.
Overview
Two small fixes for things I ran into while testing the analysis changes.
download_osm_extract.pyto adjust for that.undefined(which is what it does if the field is never set) rather thannull(which is what it ends up as if you set then clear it).Notes
I'm putting this up as a PR for visibility, but these were in place for all the testing I just did, and I don't think we need another round of testing for them, so I'm just going to merge this.