test: add coaster name matching test suite

Extract matching logic into lib/coaster-match.ts (isCoasterMatch + normalizeForMatch)
so it can be imported by both the scraper and tests without duplication.

Add tests/coaster-matching.test.ts covering all known match/false-positive cases:
- Trademark symbols, leading THE, possessives, punctuation
- Subtitle variants in both directions (Apocalypse, New Revolution - Classic)
- Space-split brand words (BAT GIRL vs Batgirl)
- 4D subtitle extension (THE JOKER™ 4D Free Fly Coaster vs Joker)
- False positives: Joker y Harley Quinn, conjunction connectors

Run with: npm test

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

lib/coaster-match.ts

@@ -0,0 +1,64 @@
+/**
+ * Coaster name matching — shared between the Queue-Times scraper and tests.
+ *
+ * Queue-Times and RCDB use different name conventions:
+ *   - Trademark symbols (™ ® ©)
+ *   - Leading "THE " prefixes
+ *   - Possessives ("Catwoman's" vs "Catwoman")
+ *   - Subtitles added or dropped ("Apocalypse" vs "Apocalypse the Ride")
+ *   - Space-split brand words ("BAT GIRL" vs "Batgirl")
+ *   - Conjunction-joined compound rides ("Joker y Harley Quinn" ≠ "Joker")
+ */
+
+// Words that join two ride names rather than extend one subtitle.
+// When a prefix match is found and the next word is one of these,
+// the longer name is a *different* ride, not a subtitle.
+const CONJUNCTIONS = new Set(["y", "and", "&", "with", "de", "del", "e", "et"]);
+
+/**
+ * Normalize a ride name for matching.
+ * Both sides (Queue-Times and RCDB) must be normalized with this function
+ * before any comparison so the transforms are symmetric.
+ */
+export function normalizeForMatch(name: string): string {
+  return name
+    .replace(/[\u2122\u00ae\u00a9™®©]/g, "")   // strip ™ ® ©
+    .replace(/^the\s+/i, "")                    // strip leading "THE "
+    .replace(/['\u2019]s\b/gi, "")              // strip possessives ('s / 's)
+    .replace(/[^\w\s]/g, " ")                   // all remaining punctuation → space
+    .replace(/\s+/g, " ")
+    .toLowerCase()
+    .trim();
+}
+
+/**
+ * Returns true when the Queue-Times ride name matches an entry in the RCDB
+ * coaster set (which must be built with normalizeForMatch).
+ *
+ * Matching strategy (in order):
+ * 1. Exact normalized match.
+ * 2. Compact (space-stripped) match — catches "BAT GIRL" vs "Batgirl".
+ * 3. Prefix match — the shorter normalized name is a prefix of the longer,
+ *    minimum 5 chars, unless the next word after the prefix is a conjunction
+ *    (which signals a compound ride name, not a subtitle).
+ */
+export function isCoasterMatch(qtName: string, coasterSet: Set<string>): boolean {
+  const norm = normalizeForMatch(qtName);
+  if (coasterSet.has(norm)) return true;
+
+  const compact = norm.replace(/\s/g, "");
+  for (const c of coasterSet) {
+    // Compact comparison
+    if (compact.length >= 5 && c.replace(/\s/g, "") === compact) return true;
+
+    // Prefix comparison
+    const shorter = norm.length <= c.length ? norm : c;
+    const longer  = norm.length <= c.length ? c : norm;
+    if (shorter.length >= 5 && longer.startsWith(shorter)) {
+      const nextWord = longer.slice(shorter.length).trim().split(" ")[0];
+      if (!CONJUNCTIONS.has(nextWord)) return true;
+    }
+  }
+
+  return false;
+}

lib/park-meta.ts

@@ -65,15 +65,8 @@ export function areCoastersStale(entry: ParkMeta): boolean {
  *   4. Collapse runs of whitespace
  *   5. Lowercase and trim
  */
-export function normalizeRideName(name: string): string {
+export { normalizeForMatch as normalizeRideName } from "./coaster-match.ts";
-  return name
+import { normalizeForMatch } from "./coaster-match.ts";
-    .replace(/[\u2122\u00ae\u00a9™®©]/g, "")
-    .replace(/^the\s+/i, "")
-    .replace(/[^\w\s]/g, " ")
-    .replace(/\s+/g, " ")
-    .toLowerCase()
-    .trim();
-}
 
 /**
  * Returns a Set of normalized coaster names for fast membership checks.
@@ -82,5 +75,5 @@ export function normalizeRideName(name: string): string {
 export function getCoasterSet(parkId: string, meta: ParkMetaMap): Set<string> | null {
   const entry = meta[parkId];
   if (!entry || entry.coasters.length === 0) return null;
-  return new Set(entry.coasters.map(normalizeRideName));
+  return new Set(entry.coasters.map(normalizeForMatch));
 }

lib/scrapers/queuetimes.ts

@@ -9,45 +9,7 @@
 
 const BASE = "https://queue-times.com/parks";
 
-/**
+import { isCoasterMatch, normalizeForMatch } from "../coaster-match.ts";
- * Normalize a ride name for fuzzy matching between Queue-Times and RCDB.
- *
- * - Strips trademark/copyright symbols (™ ® © and Unicode variants)
- * - Strips leading "THE " prefix
- * - Replaces ALL non-word, non-space characters with a space
- *   (handles !, -, :, ', ' U+2019, ", and any other punctuation)
- * - Collapses whitespace, lowercases, trims
- */
-function normalize(name: string): string {
-  return name
-    .replace(/[\u2122\u00ae\u00a9™®©]/g, "")
-    .replace(/^the\s+/i, "")
-    .replace(/[^\w\s]/g, " ")
-    .replace(/\s+/g, " ")
-    .toLowerCase()
-    .trim();
-}
-
-/**
- * Check if a Queue-Times ride name matches any coaster in the RCDB set.
- *
- * Exact normalized match covers most cases. Prefix matching handles cases
- * where one source drops or adds a subtitle:
- *   "Apocalypse" (QT) vs "Apocalypse the Ride" (RCDB)
- *   "The New Revolution - Classic" (QT) vs "New Revolution" (RCDB)
- *
- * Minimum 5 chars on the shorter side prevents accidental short matches.
- */
-function isCoaster(name: string, coasterSet: Set<string>): boolean {
-  const norm = normalize(name);
-  if (coasterSet.has(norm)) return true;
-  for (const c of coasterSet) {
-    const shorter = norm.length <= c.length ? norm : c;
-    const longer  = norm.length <= c.length ? c : norm;
-    if (shorter.length >= 5 && longer.startsWith(shorter)) return true;
-  }
-  return false;
-}
 
 const HEADERS = {
   "User-Agent":
@@ -131,7 +93,7 @@ export async function fetchLiveRides(
           isOpen: r.is_open,
           waitMinutes: r.wait_time ?? 0,
           lastUpdated: r.last_updated,
-          isCoaster: coasterNames ? isCoaster(r.name, coasterNames) : false,
+          isCoaster: coasterNames ? isCoasterMatch(r.name, coasterNames) : false,
         });
       }
     }

package.json

@@ -10,7 +10,8 @@
     "scrape": "tsx scripts/scrape.ts",
     "scrape:force": "tsx scripts/scrape.ts --rescrape",
     "discover": "tsx scripts/discover.ts",
-    "debug": "tsx scripts/debug.ts"
+    "debug": "tsx scripts/debug.ts",
+    "test": "tsx --test tests/*.test.ts"
   },
   "dependencies": {
     "better-sqlite3": "^12.8.0",

tests/coaster-matching.test.ts

@@ -0,0 +1,90 @@
+/**
+ * Coaster name matching tests.
+ *
+ * Each case documents a real mismatch found between Queue-Times ride names
+ * and RCDB coaster names, along with the park where it was observed.
+ *
+ * Run with:  npm test
+ */
+
+import { test } from "node:test";
+import assert from "node:assert/strict";
+import { isCoasterMatch, normalizeForMatch } from "../lib/coaster-match.ts";
+
+// ── Helper ──────────────────────────────────────────────────────────────────
+
+function makeSet(...rcdbNames: string[]): Set<string> {
+  return new Set(rcdbNames.map(normalizeForMatch));
+}
+
+// ── Should MATCH (Queue-Times name → RCDB name) ──────────────────────────────
+
+test("exact match after lowercasing", () => {
+  assert.ok(isCoasterMatch("Goliath", makeSet("Goliath")));
+});
+
+test("trademark symbol stripped — BATMAN™ The Ride (Over Georgia, Magic Mountain)", () => {
+  assert.ok(isCoasterMatch("BATMAN™ The Ride", makeSet("Batman The Ride")));
+});
+
+test("leading THE stripped — THE RIDDLER Mindbender (Over Georgia)", () => {
+  assert.ok(isCoasterMatch("THE RIDDLER Mindbender", makeSet("Riddler Mindbender")));
+});
+
+test("trademark + leading THE — THE RIDDLER™'s Revenge (Magic Mountain)", () => {
+  assert.ok(isCoasterMatch("THE RIDDLER™'s Revenge", makeSet("Riddler's Revenge")));
+});
+
+test("curly apostrophe possessive stripped — CATWOMAN™ Whip (New England)", () => {
+  assert.ok(isCoasterMatch("CATWOMAN™ Whip", makeSet("Catwoman's Whip")));
+});
+
+test("straight apostrophe possessive stripped", () => {
+  assert.ok(isCoasterMatch("Riddler's Revenge", makeSet("Riddler's Revenge")));
+});
+
+test("trademark + colon punctuation — SUPERMAN™: Ultimate Flight (Over Georgia)", () => {
+  assert.ok(isCoasterMatch("SUPERMAN™: Ultimate Flight", makeSet("Superman - Ultimate Flight")));
+});
+
+test("QT drops subtitle — Apocalypse (Magic Mountain)", () => {
+  assert.ok(isCoasterMatch("Apocalypse", makeSet("Apocalypse the Ride")));
+});
+
+test("QT adds subtitle — The New Revolution - Classic (Magic Mountain)", () => {
+  assert.ok(isCoasterMatch("The New Revolution - Classic", makeSet("New Revolution")));
+});
+
+test("QT exclamation stripped — SCREAM (Magic Mountain)", () => {
+  assert.ok(isCoasterMatch("SCREAM", makeSet("Scream!")));
+});
+
+test("space-split word — BAT GIRL™: Coaster Chase (Fiesta Texas)", () => {
+  assert.ok(isCoasterMatch("BAT GIRL™: Coaster Chase", makeSet("Batgirl Coaster Chase")));
+});
+
+test("trademark + 4D subtitle — THE JOKER™ 4D Free Fly Coaster (New England)", () => {
+  assert.ok(isCoasterMatch("THE JOKER™ 4D Free Fly Coaster", makeSet("Joker")));
+});
+
+test("Great American Scream Machine — top-level QT rides array (Over Georgia)", () => {
+  assert.ok(isCoasterMatch("The Great American Scream Machine", makeSet("Great American Scream Machine")));
+});
+
+test("THE JOKER™ Funhouse Coaster — top-level QT rides array (Over Georgia)", () => {
+  assert.ok(isCoasterMatch("THE JOKER™ Funhouse Coaster", makeSet("Joker Funhouse Coaster")));
+});
+
+// ── Should NOT MATCH (false positives) ──────────────────────────────────────
+
+test("false positive: Joker y Harley Quinn ≠ Joker (Six Flags Mexico)", () => {
+  assert.ok(!isCoasterMatch("Joker y Harley Quinn", makeSet("Joker")));
+});
+
+test("false positive: unrelated ride does not match", () => {
+  assert.ok(!isCoasterMatch("SkyScreamer", makeSet("Goliath")));
+});
+
+test("false positive: short prefix with conjunction — de connector", () => {
+  assert.ok(!isCoasterMatch("Batman de Gotham", makeSet("Batman")));
+});