A lightweight Express server that simulates the Xtream Codes API for local development and end-to-end testing. Uses @faker-js/faker with deterministic seeding so every credential pair always produces the same data.

Quick Start

# Start on port 3211
pnpm nx run xtream-mock-server:serve

# Start with file-watch (auto-restart on code changes)
pnpm nx run xtream-mock-server:serve-with-watch

# Start the mock server plus the Electron app
pnpm run serve:marketing-demo

# Start the mock server plus the browser web app
pnpm run serve:marketing-demo:web

Available Scenarios (credential pairs)

Username	Password	Scenario	Live cats	VOD cats	Series cats	Items/cat	Status
`user1`	`pass1`	default	8	8	8	40	active
`large`	`large`	large catalog	20	20	20	200	active
`stress`	`stress`	stress catalog	16	16	16	120	active
`series`	`series`	series-heavy	3	4	15	30	active
`minimal`	`minimal`	minimal (edge cases)	2	2	2	5	active
`epg`	`epg`	EPG fixture	2	1	1	3	active
`emptyvod`	`emptyvod`	empty VOD metadata	2	2	2	5	active
`marketing`	`marketing`	fictional release demo	4	4	4	curated	active
`expired`	`expired`	expired account	4	4	4	10	Expired
`inactive`	`inactive`	disabled account	4	4	4	10	Disabled

Any other credential pair is auto-generated using a hash of username:password as the faker seed (6 categories, 30 items each, active account).

API Endpoints

Direct Xtream Protocol

GET /player_api.php?action=<action>&username=<u>&password=<p>[&...]

Action	Description
(none) / `get_account_info`	User info + server info
`get_live_categories`	Live TV categories
`get_vod_categories`	VOD (movie) categories
`get_series_categories`	Series categories
`get_live_streams`	Live streams (optionally filtered by `category_id`)
`get_vod_streams`	VOD streams (optionally filtered by `category_id`)
`get_series`	Series list (optionally filtered by `category_id`)
`get_vod_info?vod_id=<id>`	Full movie details
`get_series_info?series_id=<id>`	Full series info (seasons + episodes)
`get_short_epg?stream_id=<id>[&limit=N]`	EPG listings for a live channel
`get_simple_data_table?stream_id=<id>`	Full per-channel EPG schedule
`get_simple_date_table?stream_id=<id>`	Legacy typo alias for full per-channel EPG schedule

PWA CORS Proxy Endpoint

IPTVnator's PWA routes Xtream calls through a backend proxy:

GET /xtream?url=<serverUrl>&action=<action>&username=<u>&password=<p>

Response: { payload: <data>, action: <action> }

Stream URLs (stub redirects)

GET /live/<username>/<password>/<streamId>.m3u8  → HLS test stream
GET /movie/<username>/<password>/<streamId>.<ext> → HLS test stream
GET /series/<username>/<password>/<streamId>.<ext> → HLS test stream

Utility Endpoints

GET  /health   → { status: "ok", server: "xtream-mock-server", port: 3211 }
POST /reset    → clears all in-memory caches; data regenerates on next request

Example Requests

# Account info (direct)
curl "http://localhost:3211/player_api.php?username=user1&password=pass1"

# Live categories (direct)
curl "http://localhost:3211/player_api.php?username=user1&password=pass1&action=get_live_categories"

# VOD details (direct)
curl "http://localhost:3211/player_api.php?username=user1&password=pass1&action=get_vod_info&vod_id=20000"

# Series info (direct)
curl "http://localhost:3211/player_api.php?username=user1&password=pass1&action=get_series_info&series_id=30000"

# EPG for stream (direct)
curl "http://localhost:3211/player_api.php?username=user1&password=pass1&action=get_short_epg&stream_id=10000"

# Full EPG schedule (direct)
curl "http://localhost:3211/player_api.php?username=epg&password=epg&action=get_simple_data_table&stream_id=10000"

# Via PWA proxy
curl "http://localhost:3211/xtream?url=http://localhost:3211&username=user1&password=pass1&action=get_live_categories"

Playwright Integration

The mock server starts automatically with nx e2e web-e2e. Run only Xtream tests using the @xtream tag:

nx e2e web-e2e --grep "@xtream"

Test files: apps/web-e2e/src/xtream.e2e.ts

Electron Xtream EPG coverage lives in apps/electron-backend-e2e/src/xtream-epg.e2e.ts.

The Playwright tests use page.route() to redirect the app's backend proxy calls (localhost:3000/xtream**) to the mock server without modifying any application code.

Data Characteristics

Deterministic: Same credentials → same data every time (seeded faker)
Cached per session: Data generated once on first request, reused until /reset
EPG: Titles and descriptions are base64-encoded (matches real Xtream API)
Dedicated EPG fixture: epg:epg returns stable live channels plus deterministic get_short_epg and get_simple_data_table payloads for timezone-focused tests
Release screenshot fixture: marketing:marketing returns fictional live, VOD, and series data with local generated artwork under apps/xtream-mock-server/public/marketing
Timestamp precedence coverage: The epg:epg scenario intentionally shifts raw start / end strings away from start_timestamp / stop_timestamp so UI tests can prove timestamps drive rendering
Stream IDs: Live 10,000+, VOD 20,000+, Series 30,000+
Category IDs: Live 101+, VOD 201+, Series 301+

Release Demo Artwork

The marketing:marketing fixture uses 30 original fictional titles for release screenshots. Poster and backdrop PNGs are served from:

apps/xtream-mock-server/public/marketing/{poster,backdrop}/

Generate or validate those assets with:

pnpm release:artwork:dry-run
pnpm release:artwork:generate
pnpm release:artwork:validate

release:artwork:generate uses gpt-image-2 through the OpenAI Image API and requires OPENAI_API_KEY. The screenshot capture workflow only reads committed local assets; it does not call OpenAI. If a local PNG is missing, the mock server falls back to its deterministic SVG renderer for development continuity. The prompt manifest deliberately varies genres and visual media across titles so the catalog does not collapse into one superhero/poster style.

Architecture

See docs/architecture/xtream-mock-server.md for a full description of the data pipeline, response shapes, and extension points.