StealThis .dev
Pages Hard

Science — Dataset Explorer

A client-side research dataset portal with a live search bar, faceted filter rail for domain, format, license, size and release year, plus a sortable, switchable grid and list of dataset cards. Each card carries a license badge, file count, size and citation count, and opens a detail drawer exposing the variable schema, a sample data preview, version history and ready-to-copy citation and DOI actions, all over a seeded in-script catalog of fictional records.

Open in Lab
html css vanilla-js
Targets: JS HTML

Code

Dataset Explorer

A self-contained research data portal modelled on repositories like a federated open-data commons. A sticky search bar matches across titles, authors, abstracts and DOIs, while a faceted filter rail narrows results by domain, file format, license, dataset size band and a dual-handle release-year range. Active facets surface as removable chips, results update live with an accessible count, and a sort dropdown reorders by relevance, citations, recency, size or title. A grid/list view toggle restyles the same card set without re-querying.

Each dataset card shows the domain tag, title, authors and institution, a trimmed abstract, key metrics (format, size, file count, citation count) and a license badge, plus an inline download button. Selecting a card opens a focus-trapped detail drawer presenting the full abstract, an at-a-glance stat panel, a scrollable variables table (name, dtype, unit, description), a monospaced sample preview, a version history with the current release marked, and a formatted citation block.

The drawer’s actions — download archive, copy citation, copy DOI — run entirely client-side with a small toast helper and a clipboard fallback. Everything is driven by a seeded in-script array of records, so the explorer works offline with no backend, build step or external libraries. Numbers, units and citations are rendered in JetBrains Mono, prose in Source Serif 4 and controls in Inter, against the clean white-and-deep-blue institutional palette.

Illustrative UI only — fictional authors, data, and figures; not real scientific results.