Closed mbjones closed 4 years ago
Thanks @mbjones . This looks good to me.
Yeah, I think the manual edits to the eml-2.2.0.json is a reasonable way to go at this stage. The R script for generating that is a bit out of step anyway -- I had started switching it over to the tabular form, which doesn't right now seem to be indicating attribute vs element status anyway and probably should be done on a branch. I do at some point still harbor hopes of improving the schema parsing so that we can generate better first-class functions automatically (see https://github.com/cboettig/build.eml), but given that the parsing is imperfect fixing the eml-2.2.-.json
object does seem like it should be easier to maintain and verify.
Fixes the 3 validation errors described in #48.
This set of changes relies on a change to the json file in data-raw that was manually modified. So we should discuss whether there is a parsing error as well that needs to be fixed in parsing the eml schema documents.