⚠️ DEPRECATED — Use @schema DSL Instead

@std/schema is deprecated. Use the built-in @schema { ... } DSL syntax, which is shorter, declarative, and integrates directly with the query DSL.

@std/schema still works but will not receive new features. All new code should use @schema.

Before / After

// ❌ Old — @std/schema (deprecated)
let schema = import @std/schema

let UserSchema = schema.define("User", {
    id: schema.id(),
    name: schema.string({required: true, min: 1, max: 100}),
    email: schema.email({required: true}),
    age: schema.integer({min: 0, max: 150}),
    role: schema.enum("user", "admin"),
    active: schema.boolean({default: true})
})

let Users = schema.table(UserSchema, db, "users")

// ✅ New — @schema DSL
@schema User {
    id: id(auto)
    name: string(required, min: 1, max: 100)
    email: email(required, unique: true)
    age: integer(min: 0, max: 150)
    role: enum["user", "admin"] = "user"
    active: boolean = true
}

let Users = db.bind(User, "users")

Migration Reference

Type Factories → DSL Types

@std/schema (old) @schema DSL (new)
schema.string({...}) string / string(...)
schema.email({...}) email / email(...)
schema.url({...}) url / url(...)
schema.phone({...}) phone / phone(...)
schema.integer({...}) integer / int / int(...)
schema.number({...}) float / number
schema.boolean({...}) boolean / bool
schema.enum("a", "b") enum["a", "b"]
schema.date({...}) date
schema.datetime({...}) datetime
schema.money({...}) money
schema.id() id(auto)
schema.array({...}) (use typed arrays)
schema.object({...}) (use nested schemas)

Operations → DSL Equivalents

@std/schema (old) @schema DSL (new)
schema.define("User", { ... }) @schema User { ... }
schema.table(UserSchema, db, "users") db.bind(User, "users")
UserSchema.validate(data) User(data).validate()
{required: true} option (required) constraint or non-? type
{default: value} option = value after the type

New DSL Features (no @std/schema equivalent)

Documentation

See Also