Conversation
Upgrade tests to include native validation for dates and date-times and IRI's. Adds all "base" classes (ie do not reference other classes) examples.
jbrown-xentity
left a comment
jbrown-xentity
left a comment
There was a problem hiding this comment.
Please note this change looks weird, promise it will make sense in the future.
| "spatial": { | ||
| "@id": "https://example.gov/locations/conus", | ||
| "@type": "Location", | ||
| "prefLabel": "Continental United States", | ||
| "altLabel": "CONUS", | ||
| "identifier": "https://www.wikidata.org/wiki/Q578170", | ||
| "bbox": "POLYGON((-124.848974 24.396308, -66.885444 24.396308, -66.885444 49.384358, -124.848974 49.384358, -124.848974 24.396308))" | ||
| }, |
There was a problem hiding this comment.
Please note this was intentional, with implementation of IRI checking (see PR description) and the problems that will be addressed in #117, it seemed better to take this out than provide a "passing" bad example. I will make sure this and other previously created examples are all re-examined for completeness in the next PR.
zopalmer14
left a comment
zopalmer14
left a comment
There was a problem hiding this comment.
LGTM. Exciting to see the examples start to come together. Also, nice catch with the format_checker thing.
neilmb
left a comment
neilmb
left a comment
There was a problem hiding this comment.
These are great, can't wait to see more of these for the compound classes. Just a couple questions for my own edification.
| @@ -0,0 +1,3 @@ | |||
| { | |||
| "@type": "Address" | |||
There was a problem hiding this comment.
question (non-blocking): There are no required properties for Address so this example could be even more (trivially) minimal. Is just keeping the @type helpful or misleading here?
| "@id": "https://example.gov/relationships/dataset-001-data-provider", | ||
| "@type": "Relationship", | ||
| "hadRole": "dataProvider", | ||
| "relation": "https://example.gov/organizations/national-weather-service" |
There was a problem hiding this comment.
suggestion (non-blocking): If relation needs to be an IRI like it is here, then I think we should have a bad example where it doesn't have the right format and if it doesn't need to be an IRI, I think it might be nice to have another good example where it's just a free-text string.
3 participants
This adds example use cases for #33. This doesn't have all the examples, but does work out the "base" examples that don't reference other classes. Final resolution of that ticket will have to wait for second pull request.
This also raised a bug that the validator wasn't actually testing the date, date-time, and IRI format that was used in the documentation. That now occurs, and resulted in some other necessary cleanup in old examples and some code and installation changes.
Add new test cases and run through tests.
Upgrade tests to include native validation for dates and date-times and IRI's.
Adds all "base" classes (ie do not reference other classes) examples.