capri

What is capri?

Capri provides type-safe, atomic Gleam bindings for Khepri, a modern distributed database built for the BEAM virtual machine. It replaces the active record pattern with a repository approach, allowing developers to define persistence layers that match traditional domain-driven design principles. The library handles path resolution, type decoding, and validation automatically.

The package includes built-in support for schema and data migrations. These migrations are atomic, versioned, and reversible, requiring each step to increment the repository version by exactly one. Decoders are attached to every version so schemas evolve in lockstep with the data. Developers can also create unsafe migrations, though the documentation strongly advises against them.

Capri manages derived views through projections, which store typed data in local ETS tables as ephemeral caches. It supports both set projections for single values and bag projections for multiple records sharing a key. The library also provides functions to join and reset nodes within a Khepri cluster, ensuring consistent state distribution across distributed environments.

capri features

• Type-safe atomic bindings for Khepri database operations
• Repository pattern implementation for structured persistence
• Atomic, versioned, and reversible data migrations
• Typed projections cached in local ETS tables
• Cluster joining and local state reset utilities
• Automatic path validation and type decoding

Who uses capri and for what

• Building distributed BEAM applications with type-safe data access
• Managing schema evolution through reversible database migrations
• Creating ephemeral cached views using typed projections
• Implementing repository patterns without active record overhead
• Synchronizing local node state across a Khepri cluster

capri FAQ

Comments (0)

Comments and upvotes open when this product launches on ProductPerch.