Beagle SCM: A Source Code Management System That Stores AST Trees

What Beagle SCM Does

Beagle is a source code management system that stores AST (abstract syntax tree) structures rather than binary blobs. The system functions as a database hub for code-related activities, storing not just source files but also related data like tickets and CI results.

Technical Implementation

The system uses BASON (binary JSON) as its data format, which is described as "CRDT-ish" for conflict-free merging. For storage, Beagle supports virtually any key-value database, with RocksDB being the current implementation.

Current Status and Usage

Beagle is in early experimental stage and hosts itself. The documentation explicitly states: "The rest is not guaranteed. Use at your own risk."

Practical Usage Examples

From the README, here are specific commands and workflows:

• Initial project posting: $ be post //replicated.live/@gritzko/librdx
• Local storage location: $HOME/.be when no protocol specified
• Inspecting RocksDB: $ ls $HOME/.be/replicated.live/
• Checking repository state: $ cat .be shows details like //replicated.live/@gritzko/librdx
• Getting a repository: $ be get //replicated.live/@gritzko/librdx

The be repo command displays structured information including:

• Repository: replicated.live
• Project path: /@gritzko/librdx
• Branches: *main
• Base files count: 574
• Waypoints: 0

File Structure and Components

The codebase includes multiple C files for different components:

• BE.c, BE.h - Core functionality
• BE.cli.c - Command-line interface
• BE.http.c - HTTP server
• BESRV.c, BESRV.h - Server components
• BESYNC.c, BESYNC.h - Synchronization
• GREP.c, IGNO.c, VER.c - Additional utilities

The project includes documentation files covering CLI usage, HTTP interface, storage model, and design rationale.