Data Story: California State Parks Data

[feep]

Three stories:

Generate and use dataset

• Generate csv from https://www.parks.ca.gov/ParkIndex
• Generate sample file, .kmx location file to be used with gps software

Improve dataset

• meta.json
• transform.star to do what the python script did in the first story

Validate with schema.json

• structure.json

[feep]

Blockers

2.

• Invalid escape Starlark ”Invalid escape sequence \[” qri#1314

3.

• Can’t figure out how to allow None in structure.json
• Allow None from transform.star
  • Segfault qri save segfaults with transform qri#1207
• Totally broken qri, can’t use desktop or checkout the correct dataset

[feep]

We can cut the validation bit, and I can publish the dataset to a new user if you want the story to go asap.

[feep]

structure.json bit deferred to a separate story, so we can post this.

Especially since @Arqu is working on jsonschema.

qri-io/jsonschema#63 and future PRs.

[Arqu]

Oh, just to add to this, think you should definitely proceed with this and "unblock" yourself for the time being by doing what Holger did and you can refer to this for a bit more input.

In a nutshell, for the problematic onces you should do an anyOf with your current validation rules + type [string, null] to account for the None/Null values. A bit of fiddling should get you to a spot that works. It's not pretty, but should get you over the hump. The jsonschema work will take a while to land, especially the part with nullables.

[feep]

Right, I got the working JSON from Holger.

=]

Thanks, Holger.

Anyway still deferring structure.json for other reasons.

Will write later. Maybe with new "AllowNullable" Maybe without.