We should be careful not to allow whitespace to sneak into the CSVs and jsonlds. The only delimiter should be the ,; No optional whitespace.
For example, at some point recently the label for status:a became "Available ":
a,-,Available ,AVAILABLE,true,"for Sierra code ""-"", if hold present or if due date present, then requestable=false"
(From https://raw.githubusercontent.com/NYPL/nypl-core/master/vocabularies/csv/statuses.csv )
That, in turn, impacted the prefLabel: https://github.com/NYPL/nypl-core/blob/master/vocabularies/json-ld/statuses.json#L40 , which then drifted into serialized data. Which is not terrible but it breaks tests and could have other consequences downstream.
It's simple enough to trim values as a consumer of nypl-core, but ideally we'd trim it at the source so that integrators don't have to look for it.
We should be careful not to allow whitespace to sneak into the CSVs and jsonlds. The only delimiter should be the
,; No optional whitespace.For example, at some point recently the label for
status:abecame "Available ":(From https://raw.githubusercontent.com/NYPL/nypl-core/master/vocabularies/csv/statuses.csv )
That, in turn, impacted the
prefLabel: https://github.com/NYPL/nypl-core/blob/master/vocabularies/json-ld/statuses.json#L40 , which then drifted into serialized data. Which is not terrible but it breaks tests and could have other consequences downstream.It's simple enough to trim values as a consumer of nypl-core, but ideally we'd trim it at the source so that integrators don't have to look for it.