Skip to content

v3.0.0

Latest
Latest

Choose a tag to compare

View all tags
@TheMyxa TheMyxa released this 12 Nov 10:59
· 1 commit to master since this release
3.0.0
50b21a7

🚀 Telegram Parser v3.0.0 — Major Update

We're excited to announce Telegram Parser v3.0.0, a major release focused on performance, full-text search, SQLite integration, improved analytics, and smarter LLM processing.

This version introduces a new SQLite-powered architecture, making it faster and easier to explore large Telegram datasets, search across exported conversations, and analyze messages.

✨ Highlights

  • 🗄️ SQLite Integration — Persistent local indexing for exported posts, comments, and analytics.

  • 🔎 Full-Text Search (FTS5) — Fast keyword searches across Telegram comments.

  • 🌍 Global Search — Search across all exported conversations from a single interface.

  • 📊 Improved Dashboard — Faster charts, statistics, filtering, and pagination.

  • 🔥 Activity Heatmap — Visualize user activity by day of the week and hour.

  • 🤖 LLM Analysis Caching — Avoid redundant AI processing and reuse previously generated results.

  • 🔌 Enhanced MCP Server — New tools for global comment searches and retrieving AI analysis results.

  • 🐳 Improved Docker Setup — Simplified first-time configuration.

🆕 What's New

🗄️ 1. SQLite-Powered Data Storage

Introduced a local SQLite database (data/state/tg.sqlite) to efficiently index and manage exported Telegram data.

  • Persistent indexing of posts and comments.

  • Faster access to large datasets.

  • Efficient filtering, sorting, and pagination.

  • Automatic indexing of JSON exports when needed.

  • Backward compatibility with existing JSON exports.

🔎 2. Full-Text & Global Search

Search functionality has been significantly expanded.

  • SQLite FTS5-powered full-text comment search.

  • Search across multiple Telegram exports.

  • New CLI command: tg search.

  • New REST API endpoint: GET /api/search.

  • New MCP tool: search_all_comments.

📊 3. Faster Dashboard & Advanced Analytics

The Dashboard now uses SQLite-backed queries instead of repeatedly processing entire JSON files.

Improvements include:

  • Faster chart generation and statistics.

  • Server-side pagination using limit and offset.

  • Efficient message and comment filtering.

  • Filter-aware summary statistics.

  • New activity heatmap showing user engagement patterns.

🤖 4. Smarter LLM Analysis

LLM-powered analysis is now more efficient and persistent.

  • SQLite-based caching of analysis results.

  • Previously analyzed unchanged messages can be skipped.

  • Incremental saving of generated analysis.

  • Reduced redundant LLM requests.

  • Easier retrieval of stored analysis results.

💻 5. New CLI Features

tg index

Create or update SQLite indexes from existing JSON exports without fetching messages from Telegram again.

tg search

Search comments within a specific export or across all indexed exports.

🔌 6. MCP Server Enhancements

New MCP tools expand integration capabilities for AI assistants and agents:

  • search_all_comments — Search comments across indexed Telegram exports.

  • list_analysis_results — Retrieve previously generated LLM analysis results.

🌐 7. New REST API Endpoints

Endpoint | Description -- | -- GET /api/search | Global comment search GET /api/export//filtered-summary | Filter-aware statistics GET /api/export//charts | SQLite-backed chart data GET /api/export//posts | Paginated and filtered posts GET /api/export//post/ | Individual post with comments

🐳 8. Docker & Configuration Improvements

  • Automatic .env creation from .env.example on first launch.

  • Easier configuration through the Web UI.

  • Improved local state management.

  • Incremental export state stored in SQLite (export_states).

  • Legacy JSON state files remain supported for migration and fallback.

⚡ Why Upgrade?

Compared to v2.3.0, Telegram Parser v3.0.0 provides:

✅ Faster processing and navigation of large exports
✅ Powerful full-text and cross-export search
✅ More efficient Dashboard analytics
✅ Reduced redundant LLM processing
✅ Better integration with MCP-compatible AI tools
✅ Improved data persistence and indexing
✅ Simplified deployment and configuration

🔄 Compatibility

Existing JSON exports remain supported. The new SQLite index can be built from previously exported data without downloading messages again.

🔗 Links

⭐ If you find Telegram Parser useful, consider starring the repository on GitHub!

Your support helps the project grow and encourages further development.

Assets 2
Loading