Sci-Hub Telegram Bot 📚

Quoted from the Sci-Hub website: "The goal of Sci-Hub is to provide free and unrestricted access to all scientific knowledge ever published in a journal or book form."

A Telegram bot that downloads academic papers from Sci-Hub using DOI links, publisher URLs, or keyword search.

@Sci_Hubot

---

Features

🔗 Multiple Input Types

• DOI URL — https://doi.org/10.1177/193229681300700321
• Bare DOI — 10.1177/193229681300700321
• DOI prefix — DOI:10.1177/193229681300700321
• Publisher URL — https://www.nature.com/articles/laban.665
• Keyword search — /kw machine learning or /search neural networks

📥 Batch Download

Send multiple DOIs at once (max 10):

10.1177/193229681300700321, 10.3389/fsurg.2020.593367

Comma-separated, newline-separated, or multiple doi.org links.

📄 Paper Info Cards

Before downloading, see a formatted card with:

• Title, authors, journal, year, volume
• Citation count
• File size (with 50 MB limit warning)
• Abstract (truncated)
• Download button + DOI link

🔍 Keyword Search

• Powered by Semantic Scholar + CrossRef APIs
• Auto-fallback if one API is rate-limited
• Pagination with year filters
• 5 results per page

🚀 Performance

• Download queue — 5 concurrent slots, FIFO ordering
• PDF cache — repeat requests served instantly (500 MB LRU cache)
• Mirror auto-discovery — tests ~15 Sci-Hub mirrors on startup + every 30 min, sorts by speed
• Async file I/O — non-blocking cache operations

📊 Admin Dashboard

Admin-only commands (hidden from regular users):

• /stats — download counters + per-day breakdown
• /users — recent user activity
• /history — recent download log
• /ban / /unban — user management
• /status — queue status
• /mirrors — Sci-Hub mirror health

💰 Donations

• Telegram Stars ⭐ payment integration
• /donate — choose amount (25/50/100/250/500)

📢 Updates Channel

• @x0projects — updates & new bots

---

Installation

git clone git@github.com:aldimhr/scihubot.git
cd scihubot
pnpm install

Configuration

Create .env file:

BOT_TOKEN=your_telegram_bot_token

Optional (for user tracking):

SUPABASE_URL=your_supabase_url
SUPABASE_KEY=your_supabase_key

Development

# Start with auto-reload
pnpm dev

# Run tests
npx jest

# Run tests with verbose output
npx jest --verbose

Production

node bot.js

Or use the systemd service:

sudo systemctl enable scihubot
sudo systemctl start scihubot

---

Architecture

scihubot/
├── bot.js                    # Main bot setup, command registration, middleware
├── actions/
│   ├── linkEntity.js         # DOI URL / publisher URL handler
│   ├── textMessage.js        # Bare DOI / DOI prefix handler
│   ├── batchDownload.js      # Multi-DOI batch processor
│   ├── downloadCallback.js   # "Download PDF" button handler
│   ├── keyword.js            # /kw and /search commands
│   ├── donate.js             # Telegram Stars donation flow
│   ├── help.js               # /help command
│   └── command.js            # Admin commands (/stats, /ban, etc.)
├── utils/
│   ├── sciHub.js             # Sci-Hub querying with mirror discovery
│   ├── mirrorDiscovery.js    # Auto-discover + health check mirrors
│   ├── downloadFile.js       # Download + 4-layer PDF validation
│   ├── downloadQueue.js      # 5-slot concurrent queue
│   ├── cache.js              # LRU PDF cache (async I/O)
│   ├── parseDoi.js           # Multi-DOI parser (comma, newline, URLs)
│   ├── paperMeta.js          # CrossRef metadata + info card formatter
│   ├── keyword.js            # Semantic Scholar + CrossRef search
│   ├── pdfSize.js            # File size pre-check (HEAD request)
│   ├── pdfSplitter.js        # Split PDFs > 50 MB using pdf-lib
│   ├── sendPDF.js            # Shared PDF sending logic
│   ├── caption.js            # PDF caption builder
│   ├── progress.js           # Progress message helper (edits in-place)
│   ├── isPDF.js              # PDF validation (magic bytes)
│   ├── dataStore.js          # JSON-based stats/downloads/bans
│   ├── constans.js           # Config constants + messages
│   ├── index.js              # Utility barrel export
│   ├── database.js           # Optional Supabase wrapper
│   ├── citation.js           # Citation formatting
│   └── errorHandler.js       # Error logging
├── data/                     # Runtime data (gitignored)
│   ├── stats.json
│   ├── downloads.json
│   ├── bans.json
│   └── mirrors.json
├── cache/                    # PDF cache (gitignored)
├── test/
│   └── helpers.test.js       # 75 tests covering all input types
└── .env                      # Environment variables

---

Test Coverage

npx jest --verbose

Category	Tests	Coverage
DOI URL parsing	5	regex extraction, trailing slashes, complex paths
Bare DOI parsing	7	slash handling, leading slash, special chars
DOI prefix parsing	4	case-insensitive DOI:/doi:/Doi:
Publisher URL detection	3	URL classification, normalization
CrossRef metadata	4	real API calls for 3 DOIs + error handling
Info card rendering	10	title, authors, journal, abstract, size
Keyboard buttons	4	download, DOI link, channel, too-large
Keyword search	6	API fallback, pagination, year filter
PDF validation	6	valid PDF, HTML, too-small, null
File size utilities	11	bytes/KB/MB, 50 MB limit boundary
Batch DOI parsing	11	comma, newline, URLs, dedup, mixed
Entity parsing	4	url entities, text_link, non-doi URLs

Total: 75 tests

---

Tech Stack

• Runtime — Node.js
• Framework — Telegraf (Telegram Bot API)
• HTTP — axios
• PDF — pdf-lib (splitting)
• Search — Semantic Scholar API, CrossRef API
• Storage — local JSON files (optional Supabase)
• Testing — Jest

---

License

MIT

---

Support

• @x0codd — bug reports & feature requests
• @x0projects — updates & new bots