Telegram Integration#

RSSHub provides an RSS feed for public Telegram channels via the route GET /telegram/channel/:username/:routeParams? 1. The route supports two fetching strategies — web scraping (default) and Telegram MTProto API — selected at runtime based on the presence of a configured TELEGRAM_SESSION.

Route Entry Point#

• Route definition: lib/routes/telegram/channel.ts
• Radar: maps t.me/s/:username → /channel/:username 2
• Example: /telegram/channel/awesomeRSSHub
• Categories: social-media

Fetching Strategy: Web Scraping vs. MTProto API#

The handler decides the strategy on every request:

useWeb = (routeParams present) OR (TELEGRAM_SESSION not configured)

Web Scraping Path#

When useWeb is true, the handler fetches https://t.me/s/${username} (or with ?q=<searchQuery>) 3 and parses the HTML with cheerio. The URL is hardcoded to t.me — there is no configurable domain fallback for the scraping path.

A known quirk: since 2024-04-20 Telegram's t.me/s/ preview incorrectly escapes & as & inside hyperlinks with an onclick attribute; the scraper works around this inline 4.

If the channel page returns no messages and the .tgme_channel_history container is absent, the route throws a descriptive error pointing the user to verify the preview URL directly 5.

Supported message types parsed from HTML 6:
VIDEO, GIF, PHOTO, POLL, VOICE, MUSIC, DOCUMENT, LOCATION, CONTACT, STICKER, ANIMATED_STICKER, VIDEO_STICKER, plus meta-types REPLY, FORWARDED, SERVICE, VIA_BOT, UNSUPPORTED, PARTIALLY_UNSUPPORTED.

MTProto API Path#

When a TELEGRAM_SESSION is set, the handler delegates to tglib/channel.ts, which uses the telegram npm package (GramJS) to call the MTProto API directly 7. It:

• Resolves the channel peer and caches the InputPeerChannel under telegram:inputEntity:<username> 8
• Fetches the last 50 messages 9
• Groups consecutive media-only messages into a single feed item 10
• Sets item links as https://t.me/s/${username}/${message.id} and feed link as https://t.me/${username} — both hardcoded to t.me

Media in the API path is served via an internal proxy endpoint /telegram/media/:username/:messageId 12.

Configuration#

All Telegram env vars are optional 13. Without TELEGRAM_SESSION, only web scraping is available.

Config is read from lib/config.ts. MTProxy activates only when all three proxy fields are provided.

Note: Some channels (pornography, copyright, politics) are blocked on Telegram's public preview (t.me/s/). For these, only the MTProto API path (with TELEGRAM_SESSION) can access the channel.

Route Parameters (routeParams)#

Supplying routeParams forces the web scraping path (regardless of TELEGRAM_SESSION). Parameters are parsed as a query string 14. For backwards compatibility, an unrecognized value is treated as searchQuery .

Example minimal feed URL 16:

/telegram/channel/NewlearnerChannel/showLinkPreview=0&showViaBot=0&showReplyTo=0&showFwdFrom=0&showFwdFromAuthor=0&showInlineButtons=0&showMediaTagInTitle=1&showMediaTagAsEmoji=1&includeFwd=0&includeReply=1&includeServiceMsg=0&includeUnsupportedMsg=0

Key Source Files#