NHL Draft Explorer

A static web app for browsing every NHL Entry Draft pick from 1979 to present, by team and year.

Live site: https://colindaugherty25.github.io/nhl-draft-search/

Features

• Browse 48 years of NHL drafts (1979–2026)
• Per-year team dropdown with era-accurate names — 1985 lists "Hartford Whalers" and "Minnesota North Stars"; 2010 lists "Atlanta Thrashers"; 2025 lists the current 32 franchises
• Era-accurate row logos from assets.nhle.com (Whalers crest for Hartford picks, Nordiques fleur-de-lis for Quebec, etc.)
• Click any team logo (row or header) to open that franchise's NHL.com page — relocated franchises follow their lineage (HFD → Hurricanes, QUE → Avalanche, ATL Thrashers → Jets, etc.)
• Click any player name to open their NHL.com profile in a new tab
• Sortable columns — text/order columns default to ascending, stats columns default to descending (so clicking PTS once shows top scorers first). Loading/no-data rows sink to the bottom regardless of direction
• Team filter survives year changes via lineage mapping (Carolina selected → switch to 1985 → auto-reselects "Hartford Whalers")

How it works

The frontend is pure static files (index.html, app.js, styles.css) consuming pre-built JSON in data/. A GitHub Action enriches the data nightly by hitting the NHL public API for each draft pick's career stats and writes the result to data/enriched-v3-{year}.json. No runtime server, no CORS proxy, no backend in production.

Run it locally

python3 server.py

Then open http://localhost:8000. The local server serves the same static files. If you click a year whose JSON isn't yet in data/, it transparently runs the enrichment for that year and writes the result so the next click is instant.

To force a rebuild of specific years:

python3 server.py --build 2025 2024 2023

Without arguments it refreshes the latest 5 years (PREWARM_COUNT). This is the exact command GitHub Actions runs.

Tests

bash tests/run.sh

Runs three layers: Python unittest for server.py pure logic, Node node:test for js/pure.mjs, and Playwright end-to-end tests against the dev server. The same suite gates the GitHub Pages deploy. See tests/README.md for details and the convention for adding tests as new features land.

Deployment & data refresh

.github/workflows/deploy.yml builds and publishes the site to GitHub Pages on three triggers:

• Every push to main (HTML/CSS/JS changes go live within ~30s)
• Daily cron at 11:00 UTC (refreshes the latest 5 years' enriched JSON so active players' season stats stay current)
• Manual workflow_dispatch (button in the Actions tab)

actions/cache@v4 carries the data/ directory across runs so old draft years' files persist without being re-enriched every day. Deploy uses actions/deploy-pages — the artifact ships directly to Pages without committing back to main.

Project layout

• index.html / styles.css / app.js — the static frontend
• js/pure.mjs — pure logic (LINEAGE, sort comparators, etc.) shared by app.js and the JS unit tests
• server.py — local dev server + --build CLI for data enrichment
• data/ — pre-built JSON consumed by the frontend (one file per year + years.json)
• tests/ — Python + JS unit + Playwright E2E test suite
• .github/workflows/deploy.yml — test gate, daily refresh, GitHub Pages deploy