Lists every artifact currently stored under
<cache>/raw/, with size, modification time and ETag metadata
when available. Both upstream artifacts (CSV directos for
non-partitioned datasets, ZIPs for yearly datasets) and CSVs
extracted from those ZIPs appear as separate rows; the sidecar
*.etag.rds files themselves are excluded from the listing
(their content surfaces in the etag and last_modified
columns).
Value
A tibble with one row per cached file, sorted by
group then dataset then file. Columns: group
(character), dataset (character), file (basename), path
(absolute), size_bytes (integer), mtime (POSIXct), etag
(character; NA when absent), last_modified (character;
NA when absent). The tibble has an attribute
total_size_bytes (numeric scalar) with the total bytes counted
toward the cache size limit. Returns a zero-row tibble with the
same schema (and total_size_bytes = 0) when the cache is empty.
Details
Extracted CSVs have no sidecar of their own — the freshness of
the parent ZIP covers them — so they appear with
etag = NA_character_ and last_modified = NA_character_.
The returned tibble carries an attribute total_size_bytes with
the sum of size_bytes, useful for comparing against the
eviction limit set via options(cvmdata.cache_max_size_mb)
(default 100 MiB). When the total exceeds 90% of the limit, the
next download triggers LRU eviction of the oldest year directories
(yearly datasets) or artifact+sidecar pairs (non-partitioned).
See cvm_cache_path() for the option's semantics.
See also
Other cache:
cvm_cache_clear(),
cvm_cache_path(),
cvm_cache_set_path()