Add a dataset to a measurement type's _source_datasets, keeping the others

_source_datasets lists every dataset that publishes rows of a type, separated by ; ("calcofi_bottle;calcofi_2022-edna"). A type two ingests share (spiciness0: calcofi_ctd-derived per CTD bin, calcofi_bottle-derived per bottle) must not be narrowed to one of them when the other re-renders and writes its literal, so each ingest merges its own key into what the registry already holds. The existing order is kept, the new key goes last, a key is never listed twice, and an empty or NA cell becomes the key alone.

Usage

merge_source_datasets(current, dataset_key)

Arguments

current: character: the registry's _source_datasets values (NA / "" allowed).
dataset_key: the dataset key to add (one string).

Value

character, the same length as current.

Examples

merge_source_datasets(c(NA, "calcofi_ctd-derived", "a;calcofi_ctd-derived"),
                      "calcofi_ctd-derived")
#> [1] "calcofi_ctd-derived"   "calcofi_ctd-derived"   "a;calcofi_ctd-derived"
# "calcofi_ctd-derived" "calcofi_ctd-derived" "a;calcofi_ctd-derived"