replace all references and skip errors #70
Conversation
| "@context": "https://project-open-data.cio.gov/v1.1/schema/catalog.jsonld", | ||
| "describedBy": "https://resources.data.gov/schemas/dcat-us/v1.1/schema/catalog.json", | ||
| "conformsTo": "https://resources.data.gov/schemas/dcat-us/v1.1/schema", | ||
| "@context": "https://resources.data.gov/schemas/dcat-us/v1.1/schema/catalog.jsonld", |
There was a problem hiding this comment.
Don't forget that the data.json files we consume from agencies are still going to have the old https://project-open-data.cio.gov/v1.1/schema identifier, so we want to make sure our test fixtures reflect that. I don't think we should be editing any of these test samples.
| if dataset.get("license") == "https://project-open-data.cio.gov/unknown-license": | ||
| dataset["license"] = "https://github.com/project-open-data/project-open-data.github.io/blob/master/unknown-license.md" | ||
| elif dataset.get("license") == "https://project-open-data.cio.gov/open-licenses": | ||
| dataset["license"] = "https://resources.data.gov/open-licenses/" |
There was a problem hiding this comment.
I know this is meant as a temporary fix but I really want to avoid this. I'm afraid if we modify the third-party metadata like this, we find ourselves on a slippery slope. By modifying the agency's metadata, we're sort of tampering with it, diminishing it's authenticity.
As open data stewards, we want to maintain the authenticity and integrity in our chain of custody of data as much as possible using reproducible and well-known methods.
|
Since project-open-data.cio.gov has been restored, there is no need for an immediate temporary fix. |
Related to datagov-deploy#1895
This PR: