Reads a release's datasets.json — the record every dataset page, ERDDAP infoUrl, and
cc_cite()'s page line are built from (calcofi4db::build_dataset_catalog(), calcofi4db >=
4.1.0) — into a data frame, one row per record. distributions, registrations, keywords,
tables, objects and coverage$years/coverage$variables arrive as list-columns (one list
per row); provider, category, attribution, links, coverage, status arrive as nested
data frames (jsonlite's simplifyVector behavior) — tidyr::unnest_wider() flattens one when
needed.
Usage
cc_datasets(
version = "latest",
what = c("datasets", "holdings", "reference"),
base_https = "https://storage.googleapis.com/calcofi-db"
)Arguments
- version
release version (default
"latest")- what
which table of the record to return:
"datasets"(default, the 16+ integrated datasets),"holdings"(datasets known but not yet in the database) or"reference"(cruises, stations, spatial layers, bathymetry — not datasets, but linked the same way)- base_https
https root of the bucket
Value
a tibble::tibble(), one row per record
Details
A release frozen before the dataset catalog (calcofi4db < 4.1.0, before 2026-09) carries no
datasets.json; this errors naming the version rather than returning an empty table, so a
caller does not mistake "no catalog yet" for "no datasets".
Examples
if (FALSE) { # \dontrun{
ds <- cc_datasets()
ds[, c("dataset_key", "dataset_name_short")]
ds$links$page # the dataset page URL, straight from the record
cc_datasets(what = "holdings")
} # }