Reviewers' Decisions on a Survey Form

A reviewer approves or rejects a submission in the Peskas Management Platform, which writes the decision to the form's surveys_flags-<asset_id> collection under the reviewer's username and to KoBoToolbox; a reviewer can also decide in KoBoToolbox directly. A pipeline rewrites that collection on every run, so it reads the decisions from both first and carries them into the flags it writes and into its validated data.

A decision is one made by any account other than the pipeline's own. Where the collection and KoBoToolbox both hold one for a submission, the later wins. KoBoToolbox is read with list_validation_statuses(), every submission at once.

Usage

review_decisions(flags = NULL, pipeline_users, asset_id = NULL, ...)

Arguments

flags

The form's current flags collection, as returned by mdb_collection_pull(), or NULL.

pipeline_users

Character. The accounts the pipeline writes statuses as.

asset_id

Character. The KoBoToolbox asset id; NULL reads the collection only.

...

Passed to list_validation_statuses(): token, or username and password, and url for a server other than eu.kobotoolbox.org.

Value

A tibble with one row per reviewed submission: submission_id (integer), validation_status, validated_at, validated_by. When KoBoToolbox cannot be read, a warning is logged and the collection's decisions are returned alone, so the run goes on. With no pipeline account name it stops with an error, so nothing is overwritten.

See also

list_validation_statuses()

Examples

if (FALSE) { # \dontrun{
conf <- read_config()
review_decisions(
  flags = mdb_collection_pull(...),
  pipeline_users = conf$ingestion$adnap$username,
  asset_id = conf$ingestion$adnap$asset_id,
  token = conf$ingestion$adnap$token
)
} # }