Moshi.jl

模式 · Pattern matching for Julia

Moshi.jl

Type-stable ADTs, exhaustive pattern matching, and Rust-style derives — built for modern Julia.

Quick StartView on GitHub

example.jl

using Moshi.Data: @data
using Moshi.Match: @match

@data Shape begin
    Circle(Float64)
    Rect(Float64, Float64)
end

area = @match shape begin
    Shape.Circle(r) => π * r^2
    Shape.Rect(w, h) => w * h
end

Three macros, one toolkit

@data

Define algebraic data types with generics, reflection, and type-stable storage.

Learn more →

@match

Exhaustive, type-stable pattern matching with guards, splats, and ADT deconstruction.

Learn more →

@derive

Rust-style trait derivation for Show, Hash, and Eq on structs and ADTs.

Learn more →

Why Moshi?

Moshi (模式, móshì) brings ML-family patterns to Julia: sum types with @data, expressive @match, and @derive for common traits. It builds on ideas from MLStyle, SumTypes, and Expronicon.

using Moshi.Data: @data
using Moshi.Match: @match
using Moshi.Derive: @derive

@data Expr begin
    Num(Int)
    Add(Self, Self)
    Mul(Self, Self)
end

@derive Expr[Show, Hash, Eq]

eval = @match expr begin
    Expr.Num(n) => n
    Expr.Add(a, b) => eval(a) + eval(b)
    Expr.Mul(a, b) => eval(a) * eval(b)
end

Get started in 30 seconds

pkg> add Moshi

Then head to the Quick Start guide or browse the API reference.