Investigate constraining cruise names against one or more vocabs

[matdon17]

The Argo metadata field DEPLOYMENT_CRUISE_ID is rarely populated, but inconsistently when it is. To improve utility of this field, a way to better constrain its contents needs to be considered. As there is no global master index of cruises, we may have to use more than one controlled vocabulary from more than one source.

[tcarval]

As proposed for the float "ending cause" (see https://github.com/nvs-vocabs/ArgoVocabs/issues/26#issuecomment-1001588655), we should use a "key:value" syntax. When available, we should provide the URI (unversal ressource identifier) of the deployement cruise. I propose two keys :

• cruise_name: the human readable name of the cruise
• uri: the machine readable cruise universal ressource identifier ( such as a doi or a persistent ID)

Example for OVIDE 2018 deployement cruise

• DEPLOYMENT_CRUISE_ID = "cruise_name:OVIDE 2018,uri:https://doi.org/10.17600/18000510"

@RomainCancouet, @fmerceur & @vincent.bernard, does that sound correct for you ?