📘 wikt-count — streaming statistics over a Wiktionary dump

A one-off counter over a Wikimedia XML dump read from stdin. It surveys Wiktionary dumps; see Data, libraries and licences.

Run

cargo build --release -p wikt-count
lbzip2 -dc $MOVA_DATA/raw/wiktionary/<db>/<date>/*-multistream*.xml*.bz2 | target/release/wikt-count <db> summary|en|uk <out-prefix>

• Every mode writes <out-prefix>.summary.json: per namespace pages, redirects, pages with a [[link]] (MediaWiki's article count), text bytes; for main-namespace articles also the year of the last edit and the last editors (top 100, distinct count, anonymous).
• en and uk also parse the language sections of the main namespace: .langs.tsv (counters per language) and .templates.tsv (templates in the focus languages); uk adds .morph.tsv and .sumcite.tsv.

Limits

Line-based parsing of the export-0.11 format (one tag per line, escaped text), without an XML parser. Wiktionary is CC BY-SA: derived counts are for measurement; see Data and model licences.

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/tools/wikt-count/