Skip to contents

Read the build-identifier log.json from a bucket prefix

Source: R/lnk_bucket_log.R
lnk_bucket_log.Rd

Sugar over lnk_bucket_get() for the most common read: parse the log.json file at the top of a versioned S3 prefix into a named list.

Usage

lnk_bucket_log(
  prefix = "https://fresh-bc.s3.us-west-2.amazonaws.com/bcfishpass"
)

Arguments

prefix

Bucket prefix as an HTTPS URL. Defaults to NGE's bcfp dump prefix.

Value

A named list with at minimum model_version, date_completed, head_sha. Function fails loud if any of these required keys are missing — the contract with the upstream dump workflow.

Details

For NGE's bcfp dump (default prefix), log.json carries the SHA the tunnel was rebuilt from, the model_version string, and the rebuild completion timestamp. Downstream consumers (csv-sync, parity drivers) use these to stamp run inputs and tie comparison rollups to a specific upstream build.

See also

lnk_bucket_get(), lnk_baseline_append()

Other bucket: lnk_bucket_get()

Examples

if (FALSE) { # \dontrun{
log <- lnk_bucket_log()
log$model_version    # e.g. "v0.7.14-125-g6e9cf1c"
log$date_completed   # e.g. "2026-05-06T04:15:41Z"
substr(log$head_sha, 1, 7)

# Pass to lnk_baseline_append() to stamp a run.
lnk_baseline_append(log, run_label = "csv-sync-20260507",
                    path = tempfile(fileext = ".csv"))
} # }