LoginSign Up
← Back to all sparks
D

dwctaxon

ANALYTICS
Velocity0.0

R package for working with Darwin Core Taxon data. Contribute to ropensci/dwctaxon development by creating an account on GitHub.

github.com/ropensci/dwctaxon ↗

A Darwin Core validator that went quiet for two years, then surfaced only to raise its R floor

darwin-coretaxonomydata-validationropenscimaintenance
◆Current state
dwctaxon edits and validates taxonomic data held in Darwin Core format, enforcing the referential rules that make a taxonomic database internally consistent. Its last real functional change was 2.0.3 in December 2023, which loosened an over-strict uniqueness requirement in column matching. The most recent entry is a development build two years later that does nothing but set a minimum R version and bump Roxygen.
◆Where it's heading
The visible arc is a package converging on correctness rather than growing. The 2.0.3 change is the most consequential: matching a reference column no longer demands that every value in it be unique, only that the matched values be — which is what makes dct_fill_col() usable on real taxonomic tables where scientificName legitimately repeats. Around it sits compliance work: an internet-connection and URL check added purely to satisfy CRAN policy, and examples reworked to restore user settings and skip deliberate errors.
◆Prediction
The 2.0.3.9001 development stamp with an R >= 4.2.0 requirement suggests a 2.0.4 release is being prepared, most likely as maintenance rather than new validation rules. The two-year gap makes any stronger claim unsupported by the feed.

◆Recent moves

  1. 8mo ago

    Development build sets R 4.2.0 floor and bumps Roxygen

    A development build that adds an R >= 4.2.0 dependency and moves to Roxygen 7.3.2. Toolchain housekeeping, not a released version, and no change to how taxonomic data is validated.

    View source ↗
  2. 2y ago

    Column matching no longer requires globally unique reference values

    Relaxes column matching so only the matched values must be unique, not every value in the reference column. That single change makes dct_fill_col() work for filling acceptedNameUsageID from scientificName on tables where some scientific names repeat — a shape that occurs constantly in real taxonomic data and previously blocked the function outright. A CRAN-mandated connectivity check on downloads arrives alongside it.

    View source ↗
  3. 3y ago

    Example cleanup and settings restoration

    Example hygiene only — deliberately failing examples are wrapped in dontrun{}, and user settings are saved and restored around each one. Visible to package checkers, not to users.

    View source ↗