feat(cctv): add Taiwan Freeway Bureau camera pack

Read ~1,800 national freeway cameras at runtime from the Freeway Bureau's
keyless MOTC 2.0 CCTV.xml table. Each published MJPEG stream maps to the
same camera server's single-JPEG path; only those exact HTTPS pairs on the
four freeway.gov.tw camera servers are registered. On by default,
CCTV_TAIWAN_ENABLED=0 skips it.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

.env.example

@@ -237,6 +237,8 @@ AISSTREAM_API_KEY=
 # CCTV_TORONTO_LIST_URL=                               # override the Toronto RESCU camera list URL
 # CCTV_LITHUANIA_ENABLED=                              # off by default; set 1 to load Via Lietuva / Eismoinfo road cameras (photo licence unconfirmed; see DATA_SOURCES.md)
 # CCTV_LITHUANIA_MAX_SOURCES=400                       # cap Lithuanian cameras; no API key required
+# CCTV_TAIWAN_ENABLED=1                                # set 0 to skip Taiwan Freeway Bureau (MOTC) road cameras
+# CCTV_TAIWAN_MAX_SOURCES=600                          # cap Taiwan freeway cameras (1-2000); no API key required
 # CCTV_MAX_SOURCES=4000                                # catalog-wide ceiling, shared round-robin across packs
 # CCTV_PREFER_AUSTIN=1                                 # prioritize Austin cameras in the registry
 # CCTV_FORCE_AUSTIN=                                   # force live default sources when a source-pack JSON is configured

DATA_SOURCES.md

@@ -110,6 +110,7 @@ The optional Local ADS-B layer and the broadcast-FM receiver use radio you recei
   **Quebec is thinly represented**, and that is upstream rather than a filter here: the province runs its own Info-Smog program, so only a handful of Quebec stations appear in ECCC's national feed. The reporting station count also varies hour to hour as stations report or fall silent.
 
 - **Via Lietuva / Eismoinfo (Lithuania).** Public road-camera catalogs and periodically updated still images from [Eismoinfo](https://eismoinfo.lt/), credited to Via Lietuva / eismoinfo.lt in the app. **Off by default.** The camera endpoints are the eismoinfo.lt website backend and are not among the services listed in Via Lietuva's [open-data statement](https://eismoinfo.lt/eismoinfo-backend/content?id=44), whose conditions are attribution and reuse on the same terms. Until the statement is confirmed to cover the camera photos, the pack loads only when an operator sets `CCTV_LITHUANIA_ENABLED=1`; with the flag unset, empty or any other value, the server makes no request to Eismoinfo. `eismoinfo-backend/camera-info-table` is joined by ID to `layer-static-features/VKR`. Without `lks=true`, the latter returns WGS84 `[latitude, longitude]`; the table's `x/y` instead use LKS-94 metres. Only the official HTTPS `/eismoinfo-backend/image-provider/camera/last?id=…` image endpoint is registered. Catalogs use the shared 15-minute cache; frames are fetched on demand and never bundled or archived. These are periodic snapshots, not continuous video. Camera headings are low-confidence estimates. No API key is required. `CCTV_LITHUANIA_MAX_SOURCES` defaults to 400.
+- **Freeway Bureau, MOTC (Taiwan).** National freeway cameras from the Freeway Bureau's keyless [CCTV static table](https://tisvcloud.freeway.gov.tw/history/motc20/CCTV.xml) in the MOTC 2.0 standard, refreshed daily and fetched at runtime (never bundled). Freeway Bureau traffic data is published under the [Open Government Data License, Taiwan (Version 1.0)](https://data.gov.tw/license), which permits reuse with attribution; the app credits the Freeway Bureau, MOTC. The table publishes each camera's MJPEG stream on one of four official camera servers (`cctvn`, `cctvn5`, `cctvc`, `cctvs` under `freeway.gov.tw`). The frame proxy relays single images, so each published stream URL is mapped to the same server's single-JPEG path for the same camera number (`/abs2mjpg/bmjpg` → `/abs2mjpg/jpg`, `/live-view/mjpg/video.cgi` → `/live-view/jpg/image.cgi`); only those exact HTTPS pairs are registered, and the still path is not itself listed in the table. These are on-demand snapshots, not continuous video. Camera headings are low-confidence estimates. No API key is required. On by default; `CCTV_TAIWAN_ENABLED=0` skips the pack, and `CCTV_TAIWAN_MAX_SOURCES` defaults to 600 (bounded to 1–2000).
 
 - **Google Maps Platform.** You supply your own API key and are bound by [Google's ToS](https://cloud.google.com/maps-platform/terms). Google Maps Content (tiles, geocodes, places) **may not be cached, stored, rehosted, or committed** — this app only ever uses it live, which is the compliant pattern. The "Google" attribution is displayed on the globe and must stay visible. Restrict your key (see [SECURITY.md](SECURITY.md)).
 - **OpenSky Network.** Its license is **non-commercial**, and operational use of the REST API in a live product can require a prior written agreement with OpenSky — even for non-profit/government use. If you deploy this commercially, contact OpenSky for your own terms. The flights layer is a toggle and runs anonymously by default.

README.md

@@ -377,7 +377,7 @@ Twenty layers and map sources. **Eighteen have a keyless path.** Some offer addi
 | 🌍 **Earthquakes**          | Global seismic activity, last 24h                                                                                                                                                                                                                                                                                                                                                   | USGS                                    | 🟢                                                                                                  |
 | ◍ **Air quality**           | Live AQHI on the published 1–10+ health-risk scale, with a Weather-panel card showing the band legend and the highest readings in view. **Canada only** — ECCC is the first registered provider | ECCC MSC GeoMet                          | 🟢                                                                                                  |
 | 🚗 **Traffic**              | Simulated vehicles on OSM roads. With TomTom, live flow speeds drive the simulation and congestion colors below ~8 km; individual vehicle positions are not live observations                                                                                                                                                                                                       | TomTom + OSM                            | 🟢 simulation · 🟡 live flow speeds                                                                 |
-| 📹 **CCTV Mesh**            | ~4,000 public cameras (shared cap) projected _into_ the 3D space — Austin · Texas (TxDOT) · California (Caltrans) · London (TfL) · Ontario (511) · Finland (Fintraffic) · British Columbia (DriveBC) · Estonia (Tallinn, Tarktee) · Delaware (DelDOT live video) · New South Wales (Live Traffic NSW) · Calgary · Norway (Statens vegvesen) · Toronto · King County (WA) · Washington (WSDOT) · Catalonia (Servei Català de Trànsit), plus opt-in New York City (NYC DOT TMC, needs your own NYC DOT agreement) and opt-in Lithuania (Via Lietuva / Eismoinfo, photo licence unconfirmed). Positions are published; poses are estimated priors **you calibrate by dragging a gizmo on the camera itself** | City APIs                               | 🟢                                                                                                  |
+| 📹 **CCTV Mesh**            | ~4,000 public cameras (shared cap) projected _into_ the 3D space — Austin · Texas (TxDOT) · California (Caltrans) · London (TfL) · Ontario (511) · Finland (Fintraffic) · British Columbia (DriveBC) · Estonia (Tallinn, Tarktee) · Delaware (DelDOT live video) · New South Wales (Live Traffic NSW) · Calgary · Norway (Statens vegvesen) · Toronto · King County (WA) · Washington (WSDOT) · Catalonia (Servei Català de Trànsit) · Taiwan (Freeway Bureau, MOTC), plus opt-in New York City (NYC DOT TMC, needs your own NYC DOT agreement) and opt-in Lithuania (Via Lietuva / Eismoinfo, photo licence unconfirmed). Positions are published; poses are estimated priors **you calibrate by dragging a gizmo on the camera itself** | City APIs                               | 🟢                                                                                                  |
 | 📷 **Mapped ALPR Cameras**  | License-plate-reader camera locations tagged by OpenStreetMap contributors, loaded one city-sized view at a time, with **SHOW NEAREST**. Locations and tags only: no plate data, no video                                                                                                                                                                                           | OpenStreetMap (incl. DeFlock mapping)   | 🟢                                                                                                  |
 | 📷 **Street Level**         | Street-level photos behind one panel, one chip per provider: coverage on the streets, image cones, an embedded viewer and **FOLLOW**. Mapillary, Google Street View, KartaView and Panoramax | Mapillary · Google · KartaView · Panoramax | 🟢 KartaView, Panoramax · 🟡 Mapillary token · 🔴 Google Maps key |
 | 📻 **Radio**                | Geolocated world radio with an **analog tuner** — drag the needle across up to 750 stations and the globe flies to each broadcaster                                                                                                                                                                                                                                                 | Radio Browser / broadcasters            | 🟢                                                                                                  |

changelog.d/taiwan-freeway-cctv.md

@@ -0,0 +1,4 @@
+- Add Taiwan's national freeway cameras from the Freeway Bureau, MOTC to CCTV:
+  ~1,800 cameras read at runtime from the official MOTC 2.0 open-data table,
+  with on-demand snapshots and attribution. No API key is required; set
+  `CCTV_TAIWAN_ENABLED=0` to skip it.

docs/CURRENT-STATE.md

@@ -1075,6 +1075,22 @@ The images are periodically updated road-condition photographs, not continuous
 video; camera poses use the existing low-confidence heading prior.
 `CCTV_LITHUANIA_MAX_SOURCES` sets its cap (default 400, bounded to 1–1000).
 
+## Taiwan CCTV source pack
+
+The CCTV catalog includes Taiwan national freeway cameras from the Freeway
+Bureau, MOTC. One keyless request reads the daily MOTC 2.0 `CCTV.xml` table
+(~1,800 cameras); the request refuses redirects, has a 15-second deadline and a
+6 MiB response cap, and uses the shared 15-minute cache and fair catalog
+allocation. An unavailable source contributes an empty pack. The table
+publishes MJPEG stream URLs; each is mapped to the same camera server's
+single-JPEG path for the same camera number, and only those exact HTTPS pairs
+on the four `freeway.gov.tw` camera servers are registered with the frame
+proxy. Camera keys come from the server and camera number, because published
+CCTV IDs contain Chinese place names and a few repeat across cameras. Poses use
+the low-confidence heading prior. `CCTV_TAIWAN_ENABLED=0` disables the pack;
+`CCTV_TAIWAN_MAX_SOURCES` sets its cap (default 600, bounded to 1–2000).
+Gate: `src/data/cctvTaiwan.test.mjs`.
+
 ## DriveBC CCTV source pack
 
 The CCTV catalog adds DriveBC highway cameras for British Columbia alongside the

scripts/format-scope.json

@@ -114,6 +114,8 @@
   "server/providers/cctv/catalog.js",
   "server/providers/cctv/lithuania.js",
   "src/data/cctvLithuania.test.mjs",
+  "server/providers/cctv/taiwan.js",
+  "src/data/cctvTaiwan.test.mjs",
   "server/providers/cctv/constants.js",
   "server/providers/cctv/media.js",
   "server/providers/cctv/normalize.js",

server/providers/cctv/catalog.js

@@ -5,6 +5,7 @@ import { allocateSourceCap, resolveCatalogCap } from './cap.js';
 import { loadGroundHeights, joinGroundHeights } from './groundHeights.js';
 import { cameraRegion, normalizeSourceItem } from './normalize.js';
 import { loadLithuaniaSources } from './lithuania.js';
+import { loadTaiwanSources } from './taiwan.js';
 import {
   loadAustinSourcesFromOpenData,
   loadCaltransSourcesFromOpenData,
@@ -144,6 +145,11 @@ const LIVE_PACKS = [
     enabled: () => envOptIn('CCTV_LITHUANIA_ENABLED'),
     load: loadLithuaniaSources,
   },
+  {
+    name: 'taiwan',
+    enabled: () => envEnabled('CCTV_TAIWAN_ENABLED'),
+    load: loadTaiwanSources,
+  },
 ];
 /**
  * Load CCTV sources from a local JSON file (CCTV_SOURCES_FILE env or default).

server/providers/cctv/taiwan.js

@@ -0,0 +1,249 @@
+import { CCTV_SOURCE_FETCH_TIMEOUT_MS } from './constants.js';
+import {
+  cameraDisplayCode,
+  fallbackHeadingFromId,
+  prioritizeSources,
+} from './normalize.js';
+import { readResponseTextCapped } from '../common/http.js';
+
+/** Freeway Bureau (MOTC) CCTV static table in the national MOTC 2.0 XML
+ * standard, published keyless and refreshed daily (~1,800 cameras). */
+export const TAIWAN_FREEWAY_CCTV_URL =
+  'https://tisvcloud.freeway.gov.tw/history/motc20/CCTV.xml';
+/** The whole network is ~1.2 MB. */
+export const TAIWAN_MAX_CATALOG_BYTES = 6 * 1024 * 1024;
+export const DEFAULT_TAIWAN_MAX_SOURCES = 600;
+
+// The table publishes each camera's MJPEG stream. The same camera servers
+// answer a single JPEG for the same camera ID on a sibling path, which is what
+// the frame proxy can relay; only these exact pairs are ever registered.
+const STREAM_TO_FRAME = new Map([
+  ['cctvn.freeway.gov.tw', ['/abs2mjpg/bmjpg', '/abs2mjpg/jpg']],
+  ['cctvn5.freeway.gov.tw', ['/abs2mjpg/bmjpg', '/abs2mjpg/jpg']],
+  ['cctvc.freeway.gov.tw', ['/abs2mjpg/bmjpg', '/abs2mjpg/jpg']],
+  [
+    'cctvs.freeway.gov.tw',
+    ['/live-view/mjpg/video.cgi', '/live-view/jpg/image.cgi'],
+  ],
+]);
+
+const DIRECTIONS = { N: 'N', S: 'S', E: 'E', W: 'W' };
+
+/** Prioritization anchors, north to south, so a cap keeps the whole island
+ * rather than only greater Taipei. */
+const ANCHORS = [
+  { lat: 25.033, lon: 121.5654 }, // Taipei
+  { lat: 24.9936, lon: 121.301 }, // Taoyuan
+  { lat: 24.8138, lon: 120.9675 }, // Hsinchu
+  { lat: 24.7021, lon: 121.7378 }, // Yilan
+  { lat: 24.1477, lon: 120.6736 }, // Taichung
+  { lat: 22.9999, lon: 120.227 }, // Tainan
+  { lat: 22.6273, lon: 120.3014 }, // Kaohsiung
+];
+
+/** Taiwan main island with slack for the coast; the freeway network has no
+ * offshore-island cameras. */
+export function isLikelyTaiwanCoordinate(lat, lon) {
+  return (
+    Number.isFinite(lat) &&
+    Number.isFinite(lon) &&
+    lat >= 21.8 &&
+    lat <= 25.4 &&
+    lon >= 119.9 &&
+    lon <= 122.1
+  );
+}
+
+const XML_ENTITIES = {
+  amp: '&',
+  lt: '<',
+  gt: '>',
+  quot: '"',
+  apos: "'",
+};
+
+function decodeXml(text) {
+  return String(text ?? '')
+    .replace(/&(amp|lt|gt|quot|apos);/g, (_, name) => XML_ENTITIES[name])
+    .trim();
+}
+
+function tag(block, name) {
+  const match = new RegExp(`<${name}>([^<]*)</${name}>`).exec(block);
+  return match ? decodeXml(match[1]) : '';
+}
+
+/**
+ * Parse the MOTC 2.0 `CCTVList` document into flat rows. Regex-based like the
+ * Catalonia and Tarktee parsers: the schema is flat, and only the fields used
+ * below are read.
+ *
+ * @param {string} xml
+ * @returns {Array<object>}
+ */
+export function parseTaiwanCctvXml(xml) {
+  const rows = [];
+  const blockRe = /<CCTV>([\s\S]*?)<\/CCTV>/g;
+  let match;
+  while ((match = blockRe.exec(String(xml || ''))) !== null) {
+    const block = match[1];
+    rows.push({
+      cctvId: tag(block, 'CCTVID'),
+      streamUrl: tag(block, 'VideoStreamURL'),
+      lat: Number(tag(block, 'PositionLat')),
+      lon: Number(tag(block, 'PositionLon')),
+      roadName: tag(block, 'RoadName'),
+      direction: tag(block, 'RoadDirection'),
+      sectionStart: tag(block, 'Start'),
+      sectionEnd: tag(block, 'End'),
+      mile: tag(block, 'LocationMile'),
+    });
+  }
+  return rows;
+}
+
+/**
+ * Published MJPEG stream URL -> the same camera's still, plus the host and
+ * camera number that identify it, or null.
+ *
+ * @param {string} streamUrl
+ * @returns {?{frameUrl: string, host: string, camera: string}}
+ */
+export function taiwanFrame(streamUrl) {
+  let url;
+  try {
+    url = new URL(String(streamUrl ?? ''));
+  } catch {
+    return null;
+  }
+  if (url.protocol !== 'https:' || url.username || url.password || url.port)
+    return null;
+  const paths = STREAM_TO_FRAME.get(url.hostname);
+  if (!paths || url.pathname !== paths[0]) return null;
+  const camera = url.searchParams.get('camera') || '';
+  if (!/^[A-Za-z0-9-]{1,40}$/.test(camera)) return null;
+  return {
+    // Reconstruct so extra query parameters cannot alter the request.
+    frameUrl: `https://${url.hostname}${paths[1]}?camera=${camera}`,
+    host: url.hostname.split('.')[0],
+    camera,
+  };
+}
+
+/**
+ * One parsed Freeway Bureau row -> one catalog source, or null.
+ *
+ * @param {object} row - Output of parseTaiwanCctvXml.
+ * @returns {?object}
+ */
+export function taiwanCameraToSource(row) {
+  if (!row || typeof row !== 'object') return null;
+  const { lat, lon } = row;
+  if (!isLikelyTaiwanCoordinate(lat, lon)) return null;
+  const frame = taiwanFrame(row.streamUrl);
+  if (!frame) return null;
+  const imageUrl = frame.frameUrl;
+  // CCTVID carries Chinese place names and a few IDs repeat across distinct
+  // cameras, so the stable key is the camera server plus its camera number.
+  const rawId = String(row.cctvId ?? '')
+    .trim()
+    .slice(0, 80);
+  const cameraId = `tw-freeway-${frame.host}-${frame.camera.toLowerCase()}`;
+  const road = String(row.roadName ?? '').trim();
+  const direction = DIRECTIONS[String(row.direction ?? '').trim()] || '';
+  const mile = String(row.mile ?? '').trim();
+  const section = [row.sectionStart, row.sectionEnd]
+    .map((part) => String(part ?? '').trim())
+    .filter(Boolean)
+    .join('→');
+  const name =
+    [road, direction, mile].filter(Boolean).join(' ') +
+      (section ? ` ${section}` : '') ||
+    rawId ||
+    `Freeway camera ${frame.camera}`;
+
+  return {
+    id: cameraId,
+    name: name.slice(0, 160),
+    // One island-wide category in the camera picker, like Finland and Norway.
+    city: 'Taiwan / 臺灣',
+    cityId: 'taiwan',
+    provider: 'Freeway Bureau, MOTC',
+    lat,
+    lon,
+    // The table gives the carriageway direction, not the camera's bearing,
+    // so the pose uses the shared low-confidence prior.
+    headingDeg: fallbackHeadingFromId(cameraId),
+    headingConfidence: 'low',
+    pitchDeg: -18,
+    fovDeg: 44,
+    rangeM: 145,
+    mountHeightM: 8,
+    feedType: 'image',
+    url: imageUrl,
+    snapshotUrl: imageUrl,
+    sourceKind: 'tw-freeway-motc',
+    license:
+      'Freeway Bureau, MOTC open data, Open Government Data License, Taiwan (Version 1.0)',
+    code: cameraDisplayCode(
+      `${road} ${mile}`.trim().toUpperCase() || frame.camera,
+    ),
+  };
+}
+
+/**
+ * Fetch Taiwan Freeway Bureau cameras from the keyless MOTC 2.0 XML table.
+ * One bounded request per shared catalog refresh.
+ *
+ * @returns {Promise<Array<object>>}
+ */
+export async function loadTaiwanSources() {
+  try {
+    const signal = AbortSignal.timeout(CCTV_SOURCE_FETCH_TIMEOUT_MS);
+    const response = await fetch(
+      process.env.CCTV_TAIWAN_URL || TAIWAN_FREEWAY_CCTV_URL,
+      {
+        headers: { Accept: 'application/xml,text/xml' },
+        redirect: 'manual',
+        signal,
+      },
+    );
+    if (!response.ok) {
+      await response.body?.cancel().catch(() => {});
+      console.warn(
+        '[CCTV] Taiwan camera download failed:',
+        response.status >= 300 && response.status < 400
+          ? 'catalog redirected; redirects are not followed'
+          : response.status,
+      );
+      return [];
+    }
+    const xml = await readResponseTextCapped(
+      response,
+      TAIWAN_MAX_CATALOG_BYTES,
+      signal,
+    );
+    const cameras = new Map();
+    for (const row of parseTaiwanCctvXml(xml)) {
+      const camera = taiwanCameraToSource(row);
+      if (camera && !cameras.has(camera.id)) cameras.set(camera.id, camera);
+    }
+    const rawCap = Number(
+      process.env.CCTV_TAIWAN_MAX_SOURCES || DEFAULT_TAIWAN_MAX_SOURCES,
+    );
+    const cap = Number.isFinite(rawCap)
+      ? Math.max(1, Math.min(2000, Math.floor(rawCap)))
+      : DEFAULT_TAIWAN_MAX_SOURCES;
+    const sources = prioritizeSources([...cameras.values()], cap, ANCHORS);
+    console.log(
+      `[CCTV] Loaded Taiwan camera sources: ${cameras.size} (using ${sources.length})`,
+    );
+    return sources;
+  } catch (error) {
+    console.warn(
+      '[CCTV] Taiwan camera download error:',
+      error?.message || error,
+    );
+    return [];
+  }
+}

src/data/cctvTaiwan.test.mjs

@@ -0,0 +1,200 @@
+import { test } from 'node:test';
+import assert from 'node:assert/strict';
+import {
+  loadTaiwanSources,
+  parseTaiwanCctvXml,
+  taiwanCameraToSource,
+  taiwanFrame,
+  TAIWAN_FREEWAY_CCTV_URL,
+} from '../../server/providers/cctv/taiwan.js';
+import { createCctvCatalog } from '../../server/providers/cctv/catalog.js';
+
+function cctvXml(...blocks) {
+  return `<?xml version='1.0' encoding='utf-8'?>
+<CCTVList xmlns="http://traffic.transportdata.tw/standard/traffic/schema/">
+  <AuthorityCode>NFB</AuthorityCode>
+  <CCTVs>${blocks.join('')}</CCTVs>
+</CCTVList>`;
+}
+
+function block({
+  id = 'CCTV-N1-S-0.000-M',
+  stream = 'https://cctvn.freeway.gov.tw/abs2mjpg/bmjpg?camera=10000',
+  lat = '25.1229931',
+  lon = '121.735695',
+} = {}) {
+  return `<CCTV>
+    <CCTVID>${id}</CCTVID>
+    <VideoStreamURL>${stream}</VideoStreamURL>
+    <PositionLon>${lon}</PositionLon>
+    <PositionLat>${lat}</PositionLat>
+    <RoadName>國道1號</RoadName>
+    <RoadDirection>S</RoadDirection>
+    <RoadSection><Start>基隆端</Start><End>基隆交流道</End></RoadSection>
+    <LocationMile>0K+000</LocationMile>
+  </CCTV>`;
+}
+
+test('parses the MOTC 2.0 table and registers the same camera still', () => {
+  const [row] = parseTaiwanCctvXml(cctvXml(block()));
+  const camera = taiwanCameraToSource(row);
+  assert.equal(camera.id, 'tw-freeway-cctvn-10000');
+  assert.equal(camera.lat, 25.1229931);
+  assert.equal(camera.lon, 121.735695);
+  assert.equal(camera.name, '國道1號 S 0K+000 基隆端→基隆交流道');
+  assert.equal(camera.cityId, 'taiwan');
+  assert.equal(camera.feedType, 'image');
+  assert.equal(camera.headingConfidence, 'low');
+  assert.equal(
+    camera.url,
+    'https://cctvn.freeway.gov.tw/abs2mjpg/jpg?camera=10000',
+  );
+  assert.equal(camera.snapshotUrl, camera.url);
+  assert.match(camera.license, /Open Government Data License, Taiwan/);
+});
+
+test('each camera server maps its stream path to its own still path only', () => {
+  assert.equal(
+    taiwanFrame(
+      'https://cctvs.freeway.gov.tw/live-view/mjpg/video.cgi?camera=373',
+    ).frameUrl,
+    'https://cctvs.freeway.gov.tw/live-view/jpg/image.cgi?camera=373',
+  );
+  assert.equal(
+    taiwanFrame(
+      'https://cctvn5.freeway.gov.tw/abs2mjpg/bmjpg?camera=c3a580c9-0232-46a1-bfc6-927590758ab4',
+    ).frameUrl,
+    'https://cctvn5.freeway.gov.tw/abs2mjpg/jpg?camera=c3a580c9-0232-46a1-bfc6-927590758ab4',
+  );
+  assert.equal(
+    taiwanFrame(
+      'https://cctvc.freeway.gov.tw/abs2mjpg/bmjpg?camera=495&extra=1',
+    ).frameUrl,
+    'https://cctvc.freeway.gov.tw/abs2mjpg/jpg?camera=495',
+  );
+});
+
+test('off-host, downgraded or malformed stream URLs never enter the proxy registry', () => {
+  for (const stream of [
+    'http://cctvn.freeway.gov.tw/abs2mjpg/bmjpg?camera=10000',
+    'https://cctvn.freeway.gov.tw.evil.test/abs2mjpg/bmjpg?camera=10000',
+    'https://user:[email protected]/abs2mjpg/bmjpg?camera=10000',
+    'https://cctvn.freeway.gov.tw:8443/abs2mjpg/bmjpg?camera=10000',
+    'https://cctvn.freeway.gov.tw/live-view/mjpg/video.cgi?camera=10000',
+    'https://cctvs.freeway.gov.tw/abs2mjpg/bmjpg?camera=373',
+    'https://cctvn.freeway.gov.tw/abs2mjpg/bmjpg?camera=../../admin',
+    'https://cctvs.freeway.gov.tw/live-view/mjpg/video.cgi?cacamera=373',
+    'http://127.0.0.1/private',
+    '',
+  ]) {
+    assert.equal(taiwanFrame(stream), null, stream);
+    const [row] = parseTaiwanCctvXml(cctvXml(block({ stream })));
+    assert.equal(taiwanCameraToSource(row), null, stream);
+  }
+});
+
+test('coordinates outside Taiwan, swapped or missing are rejected', () => {
+  for (const [lat, lon] of [
+    ['121.735695', '25.1229931'],
+    ['35.68', '139.76'],
+    ['', '121.7'],
+    ['NaN', '121.7'],
+  ]) {
+    const [row] = parseTaiwanCctvXml(cctvXml(block({ lat, lon })));
+    assert.equal(taiwanCameraToSource(row), null, `${lat},${lon}`);
+  }
+  assert.equal(taiwanCameraToSource(null), null);
+});
+
+test('Chinese place names in CCTVID stay in the record without shaping the key', () => {
+  const [row] = parseTaiwanCctvXml(
+    cctvXml(block({ id: 'CCTV-N1-N-55-R-中壢服務區&amp;(北)' })),
+  );
+  const camera = taiwanCameraToSource(row);
+  assert.equal(camera.id, 'tw-freeway-cctvn-10000');
+  assert.equal(row.cctvId, 'CCTV-N1-N-55-R-中壢服務區&(北)');
+});
+
+function mockCatalog(t, xml) {
+  const requested = [];
+  t.mock.method(console, 'log', () => {});
+  t.mock.method(console, 'warn', () => {});
+  t.mock.method(globalThis, 'fetch', async (url, options) => {
+    requested.push(String(url));
+    if (String(url) === TAIWAN_FREEWAY_CCTV_URL) {
+      assert.equal(options.redirect, 'manual');
+      return new Response(xml, {
+        headers: { 'content-type': 'application/xml' },
+      });
+    }
+    return Response.json([]);
+  });
+  return requested;
+}
+
+test('loader deduplicates cameras that share a camera number', async (t) => {
+  mockCatalog(
+    t,
+    cctvXml(
+      block(),
+      block({ id: 'CCTV-N1-S-0.000-M-dup' }),
+      block({
+        stream: 'https://cctvc.freeway.gov.tw/abs2mjpg/bmjpg?camera=495',
+        lat: '24.1',
+        lon: '120.6',
+      }),
+    ),
+  );
+  assert.deepEqual((await loadTaiwanSources()).map((c) => c.id).sort(), [
+    'tw-freeway-cctvc-495',
+    'tw-freeway-cctvn-10000',
+  ]);
+});
+
+test('a redirected or failed catalog contributes an empty pack', async (t) => {
+  t.mock.method(console, 'warn', () => {});
+  for (const status of [302, 503]) {
+    t.mock.method(
+      globalThis,
+      'fetch',
+      async () =>
+        new Response('', {
+          status,
+          headers: status === 302 ? { location: 'https://evil.test/' } : {},
+        }),
+    );
+    assert.deepEqual(await loadTaiwanSources(), []);
+  }
+});
+
+async function runCatalogWithTaiwanFlag(t, flag) {
+  const requested = mockCatalog(t, cctvXml(block()));
+  const saved = process.env.CCTV_TAIWAN_ENABLED;
+  try {
+    if (flag === undefined) delete process.env.CCTV_TAIWAN_ENABLED;
+    else process.env.CCTV_TAIWAN_ENABLED = flag;
+    const sources = await createCctvCatalog({ sourceRoot: '/nonexistent' })();
+    return { requested, sources };
+  } finally {
+    if (saved === undefined) delete process.env.CCTV_TAIWAN_ENABLED;
+    else process.env.CCTV_TAIWAN_ENABLED = saved;
+  }
+}
+
+test('the Taiwan pack is in the catalog by default', async (t) => {
+  const { sources } = await runCatalogWithTaiwanFlag(t, undefined);
+  assert.equal(
+    sources.find((c) => c.id === 'tw-freeway-cctvn-10000')?.cityId,
+    'taiwan',
+  );
+});
+
+test('CCTV_TAIWAN_ENABLED=0 makes no Freeway Bureau request', async (t) => {
+  const { requested, sources } = await runCatalogWithTaiwanFlag(t, '0');
+  assert.ok(requested.length > 0, 'the other packs still refreshed');
+  assert.equal(
+    requested.some((href) => href.includes('freeway.gov.tw')),
+    false,
+  );
+  assert.ok(!sources.some((c) => c.cityId === 'taiwan'));
+});

src/data/dataCredits.js

@@ -257,6 +257,13 @@ export const DATA_CREDITS = [
       'Road-camera stills (Lithuania, when enabled): ' +
       '<a href="https://eismoinfo.lt/" target="_blank" rel="noopener">Via Lietuva / Eismoinfo</a>; periodically updated images',
   },
+  {
+    key: 'taiwan-cctv',
+    html:
+      'Freeway camera stills (Taiwan): ' +
+      '<a href="https://www.freeway.gov.tw/" target="_blank" rel="noopener">Freeway Bureau, MOTC</a>, ' +
+      'Open Government Data License, Taiwan (Version 1.0)',
+  },
   {
     key: 'calgary-cctv',
     html: