The table of tokens that name what an output is: how its streams were cut
(cut), which access rules decided where gradient barriers cut them
(acc), how its gradient was measured (grad) and where its discharge came
from (flow). lnk_vocabulary_resolve() matches a config against it, and
every lnk_pipeline_run() records the four tokens in its <schema>.log
row.
Value
A data frame, one row per token, with columns token, axis
(cut, acc, grad or flow), definition (prose), spec (the
canonical string or digest lnk_vocabulary_resolve() matches),
introduced (the link version that added it) and retired (the link
version that retired it, NA while live).
Details
Tokens are lowercase, never renamed and never reused. A changed definition
is a new token, and a retired token stays in the table so old names still
decode. Consumers read this table (or inst/extdata/vocabulary.csv at a
release tag) rather than copying it.
See also
Other vocabulary:
lnk_vocabulary_resolve()
Examples
v <- lnk_vocabulary_read()
v[, c("token", "axis", "introduced")]
#> token axis introduced
#> 1 cut-full-top cut 0.70.0
#> 2 cut-full-ssf-all cut 0.70.0
#> 3 acc-bcfp acc 0.70.0
#> 4 acc-v1 acc 0.70.0
#> 5 grad-seg grad 0.70.0
#> 6 q-fwapg flow 0.70.0
#> 7 q-wet-fill flow 0.70.0
#> 8 acc-v2 acc 0.72.0
#> 9 acc-bcfp-2 acc 0.72.0
#> 10 acc-v3 acc 0.73.0
#> 11 acc-bcfp-3 acc 0.73.0
#> 12 acc-bcfp-4 acc 0.74.0
#> 13 acc-v4 acc 0.75.0
#> 14 acc-v5 acc 0.76.0
#> 15 cut-min cut 0.77.0
#> 16 grad-parent grad 0.79.0
#> 17 grad-fwa grad 0.81.0
#> 18 cut-full-top-2 cut 0.84.0
#> 19 cut-full-ssf-all-2 cut 0.84.0
#> 20 acc-v6 acc 0.84.0
#> 21 acc-bcfp-5 acc 0.84.0
