Data Cleaner

A fast, zero-dependency CSV cleaning tool that runs entirely in your browser — no install, no upload, no backend. Drop in a messy spreadsheet export, get clean, deduplicated, analysis-ready data back. Your data never leaves your machine.

🔗 Live demo: https://ghravenlabs.github.io/Data-Cleaner-Tool/

Portfolio proof

• Case study — why this tool matters, what it proves, and how it could be sold as a small data-cleanup service.
• GitHub Actions runs CSV/TSV regression tests and checks the static app, demo URL, and screenshot on every push.

Why

Messy CSV/Excel exports are everywhere — stray whitespace, duplicate rows, empty rows, inconsistent headers. This cleans them in one click, in pure client-side JavaScript, so it's safe for sensitive data (nothing is ever sent to a server).

Features

• Load a .csv/.tsv (file or paste), or try the built-in sample
• Trim whitespace · collapse double spaces · remove empty rows · remove duplicate rows
• Remove empty columns · standardize headers (lower_snake_case) · find & replace
• Live preview (before/after) with row/column counts and a summary of exactly what changed
• Export the cleaned CSV or copy to clipboard
• Robust CSV parser (handles quoted fields, embedded commas/newlines, auto-detects comma vs tab)

Tech

• Vanilla HTML/CSS/JavaScript — no frameworks, no build step, no dependencies
• Custom CSV parse + serialize (quote-aware), in-browser only
• Single self-contained file → works offline and deploys anywhere static

Delimiter detection uses the first logical row (the headers), ignoring separators inside quoted fields. Comma is the default when the header is ambiguous. Use consistent column separators; automatic detection cannot disambiguate every malformed or mixed-format file. Empty-column removal checks the widest row, so populated cells beyond a short header are retained. Missing header names and cells are exported as empty strings. The preview and column count also include columns beyond a short header, using blank header cells. The row preview is limited to 50 data rows; column counts cover the entire file. Pasting preserves leading/trailing tabs and cell whitespace; trimming is applied only when you select the trim option and clean the data.

Development checks

With Node.js 24 installed, run node --test tests/csv.test.cjs. The tests exercise the app's actual parser and exporter using Node's built-in test runner; no package install is needed.

Run it

• Locally: open index.html in any browser.
• Deploy: push to GitHub and enable Pages (Settings → Pages → deploy from master) — it's a static site, so it just works.

Privacy

100% client-side. No analytics, no network calls, no uploads. Your CSV is processed in the page and discarded when you close the tab.

License

MIT © Rolly Calma (Ghraven)