v0.1.0

First public release of datalint — a local, deterministic CLI that profiles and lints CSV/TSV data quality.

Profiles

Per-column inferred type, empty rate, distinct count, min/max/mean and top values.

Catches

• 🧱 Ragged rows, duplicate/empty headers, empty columns/rows
• 🔢 Type drift (numbers mixed with text), numeric outliers (Tukey/IQR)
• 🈳 Missing values, leading/trailing whitespace
• 📅 Mixed date formats, inconsistent casing (US vs us)
• ♻️ Duplicate rows
• 📐 Optional schema: required, type, enum, range, pattern, unique, not-null

Features

• Hand-rolled RFC 4180 CSV/TSV parser (no data-lib dependency) + delimiter auto-detection.
• Quality score + A–F grade, per file and overall.
• JSON + Markdown export; colored console output.
• --min-score CI gate (non-zero exit).
• Config file: delimiter, header, thresholds, schema, rule severities.
• Programmatic API (analyze, buildReport, parseCsv, toMarkdown). ESM + CJS + types.
• No Python, no API key, no server, nothing uploaded.

Install

npx @didrod2539/datalint scan data.csv

💖 Sponsor via Lemon Squeezy.