Special remote airports + isRemote curation (ISS-097, v0.61.0)

scripts/build-world-airports.mjs

@@ -3,6 +3,10 @@
 //
 // Pipeline: download → keep `large_airport` rows only (same density curation as the EU file,
 // v0.2.8) → greedy 40 km haversine clustering → one airport per cluster ("the biggest").
+// v0.61.0 (ISS-097): additionally keeps every non-large fixed-wing row that is REMOTE — no other
+// airport within 1000 km of the kept set (lib/airports.ts isRemote) — so Antarctica, the high
+// Arctic, Alaska, northern Canada and far islands are represented even though they carry no
+// large_airport; same-station twins collapse via the same 40 km clustering.
 // The output keeps the exact OurAirports header so lib/airports.ts parses it unchanged.
 //
 // Usage: node scripts/build-world-airports.mjs   (writes data/airports_world.csv, prints stats)
@@ -14,6 +18,22 @@ import path from 'node:path';
 const SOURCE_URL = 'https://davidmegginson.github.io/ourairports-data/airports.csv'; // legacy ourairports.com/data URL is dead (see CHANGELOG v0.1.0)
 const OUT_FILE = path.join(import.meta.dirname, '..', 'data', 'airports_world.csv');
 const CLUSTER_RADIUS_KM = 40; // mid of the issue's 30–50 km range — n≈1k rows makes O(n²) trivial
+// v0.61.0 (ISS-097): mirrors REMOTE_RADIUS_KM in lib/airports.ts — "no other airport within this
+// radius" is what makes a small airport worth keeping (Antarctica, high Arctic, far islands).
+const REMOTE_RADIUS_KM = 1000;
+
+// Same fixed-wing type set as INCLUDED_TYPES in lib/airports.ts (standalone copy — the script must
+// stay dependency-free): all fixed-wing airports plus seaplane bases.
+const FIXED_WING_TYPES = new Set(['airport', 'small_airport', 'medium_airport', 'large_airport', 'seaplane_base']);
+
+// Standalone copy of isRemote in lib/airports.ts, for CSV rows: true when `row` has no member of
+// `others` within REMOTE_RADIUS_KM. Callers pass disjoint sets (the row itself is not in `others`).
+function isRemoteRow(row, others) {
+  for (const o of others) {
+    if (haversineKm(row, o) <= REMOTE_RADIUS_KM) return false;
+  }
+  return true;
+}
 
 // --- Minimal RFC-4180 parser (same semantics as lib/airports.ts parseCsv, standalone copy) ---
 function parseCsv(text) {
@@ -86,26 +106,33 @@ const header = headerLine.replace(/\r$/, '').split(',').map((h) => h.replace(/^"
 if (header.length !== 19) { console.error(`unexpected header (${header.length} columns): ${headerLine}`); process.exit(1); }
 
 const allRows = parseCsv(text);
-const large = allRows.filter((r) => r.type === 'large_airport');
-console.log(`downloaded ${allRows.length} rows; ${large.length} are large_airport`);
+// v0.61.0 (ISS-097): curation runs over ALL fixed-wing rows, not just large_airport — the remote
+// pass below needs the full pool to test against.
+const fixed = allRows.filter((r) => FIXED_WING_TYPES.has(r.type));
+const large = fixed.filter((r) => r.type === 'large_airport');
+console.log(`downloaded ${allRows.length} rows; ${fixed.length} are fixed-wing, ${large.length} of them large_airport`);
 
 // Greedy clustering in file order: a row joins the first cluster whose anchor is within the
-// radius, otherwise it starts a new one. Deterministic — same input, same output.
-const clusters = []; // { anchorIdx (into `large`), members }
-for (let i = 0; i < large.length; i++) {
-  let joined = false;
-  for (const c of clusters) {
-    if (haversineKm(large[i], large[c.anchorIdx]) <= CLUSTER_RADIUS_KM) { c.members.push(i); joined = true; break; }
+// radius, otherwise it starts a new one. Deterministic — same input, same output. Returns the
+// "biggest" member (one airport per cluster) plus the clusters themselves for reporting.
+function clusterWinners(rows) {
+  const clusters = []; // { anchorIdx (into rows), members }
+  for (let i = 0; i < rows.length; i++) {
+    let joined = false;
+    for (const c of clusters) {
+      if (haversineKm(rows[i], rows[c.anchorIdx]) <= CLUSTER_RADIUS_KM) { c.members.push(i); joined = true; break; }
+    }
+    if (!joined) clusters.push({ anchorIdx: i, members: [i] });
   }
-  if (!joined) clusters.push({ anchorIdx: i, members: [i] });
+  const winners = clusters.map((c) => biggestOf(c.members.map((i) => rows[i])));
+  return { winners, clusters };
 }
 
-const kept = [];
+const { winners: kept, clusters: largeClusters } = clusterWinners(large);
 const multiClusters = [];
-for (const c of clusters) {
-  const winner = biggestOf(c.members.map((i) => large[i]));
-  kept.push(winner);
+for (const c of largeClusters) {
   if (c.members.length > 1) {
+    const winner = biggestOf(c.members.map((i) => large[i]));
     multiClusters.push({
       city: winner.municipality || winner.name,
       keptIcao: winner.icao_code || winner.ident,
@@ -114,17 +141,26 @@ for (const c of clusters) {
   }
 }
 
+// v0.61.0 (ISS-097): "special" remote airports — every non-large fixed-wing row with no other
+// airport within REMOTE_RADIUS_KM of the kept set joins the dataset, so empty regions get their
+// real (small) airports instead of nothing. Same-station twins (two strips <40 km apart at one
+// base) collapse via the same greedy clustering as above.
+const remote = fixed.filter((r) => r.type !== 'large_airport' && isRemoteRow(r, kept));
+const { winners: remoteKept } = clusterWinners(remote);
+
 // --- Guards (fail loud before writing anything) ---
+const finalRows = [...kept, ...remoteKept];
 const key = (r) => r.icao_code || r.ident; // the effective ICAO consumers key by (toGeoJson fallback)
-const keys = new Set(kept.map(key));
-if (keys.size !== kept.length) { console.error('duplicate ICAO in curated set — aborting'); process.exit(1); }
+const keys = new Set(finalRows.map(key));
+if (keys.size !== finalRows.length) { console.error('duplicate ICAO in curated set — aborting'); process.exit(1); }
 for (const must of ['EDDF', 'UNNT', 'LPPD']) { // EDDF is the seed hub; UNNT (Novosibirsk) + LPPD (Ponta Delgada) pin the 409-backstop test pair
   if (!keys.has(must)) { console.error(`required airport ${must} missing from curated set — aborting`); process.exit(1); }
 }
-if (kept.length < 300 || kept.length > 2500) { console.error(`unexpected row count ${kept.length}`); process.exit(1); }
+if (remoteKept.length === 0) { console.error('no remote airports found in the dump — curation rule broken, aborting'); process.exit(1); }
+if (finalRows.length < 300 || finalRows.length > 2500) { console.error(`unexpected row count ${finalRows.length}`); process.exit(1); }
 
-await fs.writeFile(OUT_FILE, [headerLine.replace(/\r$/, ''), ...kept.map((r) => header.map((h) => `"${String(r[h] ?? '').replace(/"/g, '""')}"`).join(','))].join('\n') + '\n', 'utf8');
+await fs.writeFile(OUT_FILE, [headerLine.replace(/\r$/, ''), ...finalRows.map((r) => header.map((h) => `"${String(r[h] ?? '').replace(/"/g, '""')}"`).join(','))].join('\n') + '\n', 'utf8');
 
-console.log(`wrote ${OUT_FILE}: ${kept.length} rows kept, ${large.length - kept.length} dropped into ${multiClusters.length} multi-airport clusters`);
+console.log(`wrote ${OUT_FILE}: ${kept.length} large + ${remoteKept.length} remote = ${finalRows.length} rows (${large.length - kept.length} large dropped into ${multiClusters.length} multi-airport clusters)`);
 console.log('sample merged cities (kept ← dropped):');
 for (const c of multiClusters.slice(0, 15)) console.log(`  ${c.city}: kept ${c.keptIcao} ← ${c.dropped.join(', ')}`);