Basic Usage

Primitives

The Schema module provides built-in schemas for common primitive types.

Schema	Equivalent TypeScript Type
Schema.String	string
Schema.Number	number
Schema.Boolean	boolean
Schema.BigIntFromSelf	BigInt
Schema.SymbolFromSelf	symbol
Schema.Object	object
Schema.Undefined	undefined
Schema.Void	void
Schema.Any	any
Schema.Unknown	unknown
Schema.Never	never

Example (Using a Primitive Schema)

import { Schema } from "effect"

const schema = Schema.String

// Infers the type as string
//
//     ┌─── string
//     ▼
type Type = typeof schema.Type

// Attempt to decode a null value, which will throw a parse error
Schema.decodeUnknownSync(schema)(null)
/*
throws:
ParseError: Expected string, actual null
*/

asSchema

To make it easier to work with schemas, built-in schemas are exposed with shorter, opaque types when possible.

The Schema.asSchema function allows you to view any schema as Schema<Type, Encoded, Context>.

Example (Expanding a Schema with asSchema)

For example, while Schema.String is defined as a class with a type of typeof Schema.String, using Schema.asSchema provides the schema in its extended form as Schema<string, string, never>.

import { Schema } from "effect"

//     ┌─── typeof Schema.String
//     ▼
const schema = Schema.String

//     ┌─── Schema<string, string, never>
//     ▼
const nomalized = Schema.asSchema(schema)

Unique Symbols

You can create a schema for unique symbols using Schema.UniqueSymbolFromSelf.

Example (Creating a Schema for a Unique Symbol)

import { Schema } from "effect"

const mySymbol = Symbol.for("mySymbol")

const schema = Schema.UniqueSymbolFromSelf(mySymbol)

//     ┌─── typeof mySymbol
//     ▼
type Type = typeof schema.Type

Schema.decodeUnknownSync(schema)(null)
/*
throws:
ParseError: Expected Symbol(mySymbol), actual null
*/

Literals

Literal schemas represent a literal type. You can use them to specify exact values that a type must have.

Literals can be of the following types:

• string
• number
• boolean
• null
• bigint

Example (Defining Literal Schemas)

import { Schema } from "effect"

// Define various literal schemas
Schema.Null // Same as S.Literal(null)
Schema.Literal("a") // string literal
Schema.Literal(1) // number literal
Schema.Literal(true) // boolean literal
Schema.Literal(2n) // BigInt literal

Example (Defining a Literal Schema for "a")

import { Schema } from "effect"

//     ┌─── Literal<["a"]>
//     ▼
const schema = Schema.Literal("a")

//     ┌─── "a"
//     ▼
type Type = typeof schema.Type

console.log(Schema.decodeUnknownSync(schema)("a"))
// Output: "a"

console.log(Schema.decodeUnknownSync(schema)("b"))
/*
throws:
ParseError: Expected "a", actual "b"
*/

Union of Literals

You can create a union of multiple literals by passing them as arguments to the Schema.Literal constructor:

Example (Defining a Union of Literals)

import { Schema } from "effect"

//     ┌─── Literal<["a", "b", "c"]>
//     ▼
const schema = Schema.Literal("a", "b", "c")

//     ┌─── "a" | "b" | "c"
//     ▼
type Type = typeof schema.Type

Schema.decodeUnknownSync(schema)(null)
/*
throws:
ParseError: "a" | "b" | "c"
├─ Expected "a", actual null
├─ Expected "b", actual null
└─ Expected "c", actual null
*/

If you want to set a custom error message for the entire union of literals, you can use the override: true option (see Custom Error Messages for more details) to specify a unified message.

Example (Adding a Custom Message to a Union of Literals)

import { Schema } from "effect"

// Schema with individual messages for each literal
const individualMessages = Schema.Literal("a", "b", "c")

console.log(Schema.decodeUnknownSync(individualMessages)(null))
/*
throws:
ParseError: "a" | "b" | "c"
├─ Expected "a", actual null
├─ Expected "b", actual null
└─ Expected "c", actual null
*/

// Schema with a unified custom message for all literals
const unifiedMessage = Schema.Literal("a", "b", "c").annotations({
  message: () => ({ message: "Not a valid code", override: true })
})

console.log(Schema.decodeUnknownSync(unifiedMessage)(null))
/*
throws:
ParseError: Not a valid code
*/

Exposed Values

You can access the literals defined in a literal schema using the literals property:

import { Schema } from "effect"

const schema = Schema.Literal("a", "b", "c")

//      ┌─── readonly ["a", "b", "c"]
//      ▼
const literals = schema.literals

The pickLiteral Utility

You can use Schema.pickLiteral with a literal schema to narrow down its possible values.

Example (Using pickLiteral to Narrow Values)

import { Schema } from "effect"

// Create a schema for a subset of literals ("a" and "b") from a larger set
//
//      ┌─── Literal<["a", "b"]>
//      ▼
const schema = Schema.Literal("a", "b", "c").pipe(
  Schema.pickLiteral("a", "b")
)

Sometimes, you may need to reuse a literal schema in other parts of your code. Below is an example demonstrating how to do this:

Example (Creating a Subtype from a Literal Schema)

import { Schema } from "effect"

// Define the base set of fruit categories
const FruitCategory = Schema.Literal("sweet", "citrus", "tropical")

// Define a general Fruit schema with the base category set
const Fruit = Schema.Struct({
  id: Schema.Number,
  category: FruitCategory
})

// Define a specific Fruit schema for only "sweet" and "citrus" categories
const SweetAndCitrusFruit = Schema.Struct({
  id: Schema.Number,
  category: FruitCategory.pipe(Schema.pickLiteral("sweet", "citrus"))
})

In this example, FruitCategory serves as the source of truth for the different fruit categories. We reuse it to create a subtype of Fruit called SweetAndCitrusFruit, ensuring that only the specified categories ("sweet" and "citrus") are allowed. This approach helps maintain consistency throughout your code and provides type safety if the category definition changes.

Template literals

In TypeScript, template literals types allow you to embed expressions within string literals. The Schema.TemplateLiteral constructor allows you to create a schema for these template literal types.

Example (Defining Template Literals)

import { Schema } from "effect"

// This creates a schema for: `a${string}`
//
//      ┌─── TemplateLiteral<`a${string}`>
//      ▼
const schema1 = Schema.TemplateLiteral("a", Schema.String)

// This creates a schema for:
// `https://${string}.com` | `https://${string}.net`
const schema2 = Schema.TemplateLiteral(
  "https://",
  Schema.String,
  ".",
  Schema.Literal("com", "net")
)

Example (From template literals types Documentation)

Let’s look at a more complex example. Suppose you have two sets of locale IDs for emails and footers. You can use the Schema.TemplateLiteral constructor to create a schema that combines these IDs:

import { Schema } from "effect"

const EmailLocaleIDs = Schema.Literal("welcome_email", "email_heading")
const FooterLocaleIDs = Schema.Literal("footer_title", "footer_sendoff")

// This creates a schema for:
// "welcome_email_id" | "email_heading_id" |
// "footer_title_id" | "footer_sendoff_id"
const schema = Schema.TemplateLiteral(
  Schema.Union(EmailLocaleIDs, FooterLocaleIDs),
  "_id"
)

Supported Span Types

The Schema.TemplateLiteral constructor supports the following types of spans:

• Schema.String
• Schema.Number
• Literals: string | number | boolean | null | bigint. These can be either wrapped by Schema.Literal or used directly
• Unions of the above types
• Brands of the above types

Example (Using a Branded String in a Template Literal)

import { Schema } from "effect"

// Create a branded string schema for an authorization token
const AuthorizationToken = Schema.String.pipe(
  Schema.brand("AuthorizationToken")
)

// This creates a schema for:
// `Bearer ${string & Brand<"AuthorizationToken">}`
const schema = Schema.TemplateLiteral("Bearer ", AuthorizationToken)

TemplateLiteralParser

The Schema.TemplateLiteral constructor, while useful as a simple validator, only verifies that an input conforms to a specific string pattern by converting template literal definitions into regular expressions. Similarly, Schema.pattern employs regular expressions directly for the same purpose. Post-validation, both methods require additional manual parsing to convert the validated string into a usable data format.

To address these limitations and eliminate the need for manual post-validation parsing, the Schema.TemplateLiteralParser API has been developed. It not only validates the input format but also automatically parses it into a more structured and type-safe output, specifically into a tuple format.

The Schema.TemplateLiteralParser constructor supports the same types of spans as Schema.TemplateLiteral.

Example (Using TemplateLiteralParser for Parsing and Encoding)

import { Schema } from "effect"

//      ┌─── Schema<readonly [number, "a", string], `${string}a${string}`>
//      ▼
const schema = Schema.TemplateLiteralParser(
  Schema.NumberFromString,
  "a",
  Schema.NonEmptyString
)

console.log(Schema.decodeSync(schema)("100afoo"))
// Output: [ 100, 'a', 'foo' ]

console.log(Schema.encodeSync(schema)([100, "a", "foo"]))
// Output: '100afoo'

Native enums

The Schema module provides support for native TypeScript enums. You can define a schema for an enum using Schema.Enums, allowing you to validate values that belong to the enum.

Example (Defining a Schema for an Enum)

import { Schema } from "effect"

enum Fruits {
  Apple,
  Banana
}

//      ┌─── Enums<typeof Fruits>
//      ▼
const schema = Schema.Enums(Fruits)

//
//     ┌─── Fruits
//     ▼
type Type = typeof schema.Type

Exposed Values

Enums are accessible through the enums property of the schema. You can use this property to retrieve individual members or the entire set of enum values.

import { Schema } from "effect"

enum Fruits {
  Apple,
  Banana
}

const schema = Schema.Enums(Fruits)

schema.enums // Returns all enum members
schema.enums.Apple // Access the Apple member
schema.enums.Banana // Access the Banana member

Unions

The Schema module includes a built-in Schema.Union constructor for creating “OR” types, allowing you to define schemas that can represent multiple types.

Example (Defining a Union Schema)

import { Schema } from "effect"

//      ┌─── Union<[typeof Schema.String, typeof Schema.Number]>
//      ▼
const schema = Schema.Union(Schema.String, Schema.Number)

//     ┌─── string | number
//     ▼
type Type = typeof schema.Type

Union Member Evaluation Order

When decoding, union members are evaluated in the order they are defined. If a value matches the first member, it will be decoded using that schema. If not, the decoding process moves on to the next member.

If multiple schemas could decode the same value, the order matters. Placing a more general schema before a more specific one may result in missing properties, as the first matching schema will be used.

Example (Handling Overlapping Schemas in a Union)

import { Schema } from "effect"

// Define two overlapping schemas

const Member1 = Schema.Struct({
  a: Schema.String
})

const Member2 = Schema.Struct({
  a: Schema.String,
  b: Schema.Number
})

// ❌ Define a union where Member1 appears first
const Bad = Schema.Union(Member1, Member2)

console.log(Schema.decodeUnknownSync(Bad)({ a: "a", b: 12 }))
// Output: { a: 'a' }  (Member1 matched first, so `b` was ignored)

// ✅ Define a union where Member2 appears first
const Good = Schema.Union(Member2, Member1)

console.log(Schema.decodeUnknownSync(Good)({ a: "a", b: 12 }))
// Output: { a: 'a', b: 12 } (Member2 matched first, so `b` was included)

Union of Literals

While you can create a union of literals by combining individual literal schemas:

Example (Using Individual Literal Schemas)

import { Schema } from "effect"

//      ┌─── Union<[Schema.Literal<["a"]>, Schema.Literal<["b"]>, Schema.Literal<["c"]>]>
//      ▼
const schema = Schema.Union(
  Schema.Literal("a"),
  Schema.Literal("b"),
  Schema.Literal("c")
)

You can simplify the process by passing multiple literals directly to the Schema.Literal constructor:

Example (Defining a Union of Literals)

import { Schema } from "effect"

//     ┌─── Literal<["a", "b", "c"]>
//     ▼
const schema = Schema.Literal("a", "b", "c")

//     ┌─── "a" | "b" | "c"
//     ▼
type Type = typeof schema.Type

If you want to set a custom error message for the entire union of literals, you can use the override: true option (see Custom Error Messages for more details) to specify a unified message.

Example (Adding a Custom Message to a Union of Literals)

import { Schema } from "effect"

// Schema with individual messages for each literal
const individualMessages = Schema.Literal("a", "b", "c")

console.log(Schema.decodeUnknownSync(individualMessages)(null))
/*
throws:
ParseError: "a" | "b" | "c"
├─ Expected "a", actual null
├─ Expected "b", actual null
└─ Expected "c", actual null
*/

// Schema with a unified custom message for all literals
const unifiedMessage = Schema.Literal("a", "b", "c").annotations({
  message: () => ({ message: "Not a valid code", override: true })
})

console.log(Schema.decodeUnknownSync(unifiedMessage)(null))
/*
throws:
ParseError: Not a valid code
*/

Nullables

The Schema module includes utility functions for defining schemas that allow nullable types, helping to handle values that may be null, undefined, or both.

Example (Creating Nullable Schemas)

import { Schema } from "effect"

// Represents a schema for a string or null value
Schema.NullOr(Schema.String)

// Represents a schema for a string, null, or undefined value
Schema.NullishOr(Schema.String)

// Represents a schema for a string or undefined value
Schema.UndefinedOr(Schema.String)

Discriminated unions

Discriminated unions in TypeScript are a way of modeling complex data structures that may take on different forms based on a specific set of conditions or properties. They allow you to define a type that represents multiple related shapes, where each shape is uniquely identified by a shared discriminant property.

In a discriminated union, each variant of the union has a common property, called the discriminant. The discriminant is a literal type, which means it can only have a finite set of possible values. Based on the value of the discriminant property, TypeScript can infer which variant of the union is currently in use.

Example (Defining a Discriminated Union in TypeScript)

type Circle = {
  readonly kind: "circle"
  readonly radius: number
}

type Square = {
  readonly kind: "square"
  readonly sideLength: number
}

type Shape = Circle | Square

In the Schema module, you can define a discriminated union similarly by specifying a literal field as the discriminant for each type.

Example (Defining a Discriminated Union Using Schema)

import { Schema } from "effect"

const Circle = Schema.Struct({
  kind: Schema.Literal("circle"),
  radius: Schema.Number
})

const Square = Schema.Struct({
  kind: Schema.Literal("square"),
  sideLength: Schema.Number
})

const Shape = Schema.Union(Circle, Square)

In this example, the Schema.Literal constructor sets up the kind property as the discriminant for both Circle and Square schemas. The Shape schema then represents a union of these two types, allowing TypeScript to infer the specific shape based on the kind value.

Transforming a Simple Union into a Discriminated Union

If you start with a simple union and want to transform it into a discriminated union, you can add a special property to each member. This allows TypeScript to automatically infer the correct type based on the value of the discriminant property.

Example (Initial Simple Union)

For example, let’s say you’ve defined a Shape union as a combination of Circle and Square without any special property:

import { Schema } from "effect"

const Circle = Schema.Struct({
  radius: Schema.Number
})

const Square = Schema.Struct({
  sideLength: Schema.Number
})

const Shape = Schema.Union(Circle, Square)

To make your code more manageable, you may want to transform the simple union into a discriminated union. This way, TypeScript will be able to automatically determine which member of the union you’re working with based on the value of a specific property.

To achieve this, you can add a special property to each member of the union, which will allow TypeScript to know which type it’s dealing with at runtime. Here’s how you can transform the Shape schema into another schema that represents a discriminated union:

Example (Adding Discriminant Property)

import { Schema } from "effect"

const Circle = Schema.Struct({
  radius: Schema.Number
})

const Square = Schema.Struct({
  sideLength: Schema.Number
})

const DiscriminatedShape = Schema.Union(
  Schema.transform(
    Circle,
    // Add a "kind" property with the literal value "circle" to Circle
    Schema.Struct({ ...Circle.fields, kind: Schema.Literal("circle") }),
    {
      strict: true,
      // Add the discriminant property to Circle
      decode: (circle) => ({ ...circle, kind: "circle" as const }),
      // Remove the discriminant property
      encode: ({ kind: _kind, ...rest }) => rest
    }
  ),

  Schema.transform(
    Square,
    // Add a "kind" property with the literal value "square" to Square
    Schema.Struct({ ...Square.fields, kind: Schema.Literal("square") }),
    {
      strict: true,
      // Add the discriminant property to Square
      decode: (square) => ({ ...square, kind: "square" as const }),
      // Remove the discriminant property
      encode: ({ kind: _kind, ...rest }) => rest
    }
  )
)

console.log(Schema.decodeUnknownSync(DiscriminatedShape)({ radius: 10 }))
// Output: { radius: 10, kind: 'circle' }

console.log(
  Schema.decodeUnknownSync(DiscriminatedShape)({ sideLength: 10 })
)
// Output: { sideLength: 10, kind: 'square' }

The previous solution works perfectly and shows how we can add properties to our schema at will, making it easier to consume the result within our domain model. However, it requires a lot of boilerplate. Fortunately, there is an API called Schema.attachPropertySignature designed specifically for this use case, which allows us to achieve the same result with much less effort:

Example (Using Schema.attachPropertySignature for Less Code)

import { Schema } from "effect"

const Circle = Schema.Struct({
  radius: Schema.Number
})

const Square = Schema.Struct({
  sideLength: Schema.Number
})

const DiscriminatedShape = Schema.Union(
  Circle.pipe(Schema.attachPropertySignature("kind", "circle")),
  Square.pipe(Schema.attachPropertySignature("kind", "square"))
)

// decoding
console.log(Schema.decodeUnknownSync(DiscriminatedShape)({ radius: 10 }))
// Output: { radius: 10, kind: 'circle' }

// encoding
console.log(
  Schema.encodeSync(DiscriminatedShape)({
    kind: "circle",
    radius: 10
  })
)
// Output: { radius: 10 }

Property Addition Only

Please note that with Schema.attachPropertySignature, you can only add a property, it cannot replace or override an existing one.

Exposed Values

You can access the individual members of a union schema represented as a tuple:

import { Schema } from "effect"

const schema = Schema.Union(Schema.String, Schema.Number)

// Accesses the members of the union
const members = schema.members

//      ┌─── typeof Schema.String
//      ▼
const firstMember = members[0]

//      ┌─── typeof Schema.Number
//      ▼
const secondMember = members[1]

Tuples

The Schema module allows you to define tuples, which are ordered collections of elements that may have different types. You can define tuples with required, optional, or rest elements.

Required Elements

To define a tuple with required elements, you can use the Schema.Tuple constructor and simply list the element schemas in order:

Example (Defining a Tuple with Required Elements)

import { Schema } from "effect"

// Define a tuple with a string and a number as required elements
//
//      ┌─── Tuple<[typeof Schema.String, typeof Schema.Number]>
//      ▼
const schema = Schema.Tuple(Schema.String, Schema.Number)

//     ┌─── readonly [string, number]
//     ▼
type Type = typeof schema.Type

Append a Required Element

You can append additional required elements to an existing tuple by using the spread operator:

Example (Adding an Element to an Existing Tuple)

import { Schema } from "effect"

const tuple1 = Schema.Tuple(Schema.String, Schema.Number)

// Append a boolean to the existing tuple
const tuple2 = Schema.Tuple(...tuple1.elements, Schema.Boolean)

//     ┌─── readonly [string, number, boolean]
//     ▼
type Type = typeof tuple2.Type

Optional Elements

To define an optional element, use the Schema.optionalElement constructor.

Example (Defining a Tuple with Optional Elements)

import { Schema } from "effect"

// Define a tuple with a required string and an optional number
const schema = Schema.Tuple(
  Schema.String, // required element
  Schema.optionalElement(Schema.Number) // optional element
)

//     ┌─── readonly [string, number?]
//     ▼
type Type = typeof schema.Type

Rest Element

To define a rest element, add it after the list of required or optional elements. The rest element allows the tuple to accept additional elements of a specific type.

Example (Using a Rest Element)

import { Schema } from "effect"

// Define a tuple with required elements and a rest element of type boolean
const schema = Schema.Tuple(
  [Schema.String, Schema.optionalElement(Schema.Number)], // elements
  Schema.Boolean // rest element
)

//     ┌─── readonly [string, number?, ...boolean[]]
//     ▼
type Type = typeof schema.Type

You can also include other elements after the rest:

Example (Including Additional Elements After a Rest Element)

import { Schema } from "effect"

// Define a tuple with required elements, a rest element,
// and an additional element
const schema = Schema.Tuple(
  [Schema.String, Schema.optionalElement(Schema.Number)], // elements
  Schema.Boolean, // rest element
  Schema.String // additional element
)

//     ┌─── readonly [string, number | undefined, ...boolean[], string]
//     ▼
type Type = typeof schema.Type

Annotations

Annotations are useful for adding metadata to tuple elements, making it easier to describe their purpose or requirements. This is especially helpful for generating documentation or JSON schemas.

Example (Adding Annotations to Tuple Elements)

import { JSONSchema, Schema } from "effect"

// Define a tuple representing a point with annotations for each coordinate
const Point = Schema.Tuple(
  Schema.element(Schema.Number).annotations({
    title: "X",
    description: "X coordinate"
  }),
  Schema.optionalElement(Schema.Number).annotations({
    title: "Y",
    description: "optional Y coordinate"
  })
)

// Generate a JSON Schema from the tuple
console.log(JSONSchema.make(Point))
/*
Output:
{
  '$schema': 'http://json-schema.org/draft-07/schema#',
  type: 'array',
  minItems: 1,
  items: [
    { type: 'number', description: 'X coordinate', title: 'X' },
    {
      type: 'number',
      description: 'optional Y coordinate',
      title: 'Y'
    }
  ],
  additionalItems: false
}
*/

Exposed Values

You can access the elements and rest elements of a tuple schema using the elements and rest properties:

Example (Accessing Elements and Rest Element in a Tuple Schema)

import { Schema } from "effect"

// Define a tuple with required, optional, and rest elements
const schema = Schema.Tuple(
  [Schema.String, Schema.optionalElement(Schema.Number)], // elements
  Schema.Boolean, // rest element
  Schema.String // additional element
)

// Access the required and optional elements of the tuple
//
//      ┌─── readonly [typeof Schema.String, Schema.Element<typeof Schema.Number, "?">]
//      ▼
const tupleElements = schema.elements

// Access the rest element of the tuple
//
//      ┌─── readonly [typeof Schema.Boolean, typeof Schema.String]
//      ▼
const restElement = schema.rest

Arrays

The Schema module allows you to define schemas for arrays, making it easy to validate collections of elements of a specific type.

Example (Defining an Array Schema)

import { Schema } from "effect"

// Define a schema for an array of numbers
//
//      ┌─── Array$<typeof Schema.Number>
//      ▼
const schema = Schema.Array(Schema.Number)

//     ┌─── readonly number[]
//     ▼
type Type = typeof schema.Type

Mutable Arrays

By default, Schema.Array generates a type marked as readonly. To create a schema for a mutable array, you can use the Schema.mutable function, which makes the array type mutable in a shallow manner.

Example (Creating a Mutable Array Schema)

import { Schema } from "effect"

// Define a schema for a mutable array of numbers
//
//      ┌─── mutable<Schema.Array$<typeof Schema.Number>>
//      ▼
const schema = Schema.mutable(Schema.Array(Schema.Number))

//     ┌─── number[]
//     ▼
type Type = typeof schema.Type

Exposed Values

You can access the value type of an array schema using the value property:

Example (Accessing the Value Type of an Array Schema)

import { Schema } from "effect"

const schema = Schema.Array(Schema.Number)

// Access the value type of the array schema
//
//      ┌─── typeof Schema.Number
//      ▼
const value = schema.value

Non Empty Arrays

The Schema module also provides a way to define schemas for non-empty arrays, ensuring that the array always contains at least one element.

Example (Defining a Non-Empty Array Schema)

import { Schema } from "effect"

// Define a schema for a non-empty array of numbers
//
//      ┌─── NonEmptyArray<typeof Schema.Number>
//      ▼
const schema = Schema.NonEmptyArray(Schema.Number)

//     ┌─── readonly [number, ...number[]]
//     ▼
type Type = typeof schema.Type

Exposed Values

You can access the value type of a non-empty array schema using the value property:

Example (Accessing the Value Type of a Non-Empty Array Schema)

import { Schema } from "effect"

// Define a schema for a non-empty array of numbers
const schema = Schema.NonEmptyArray(Schema.Number)

// Access the value type of the non-empty array schema
//
//      ┌─── typeof Schema.Number
//      ▼
const value = schema.value

Records

The Schema module provides support for defining record types, which are collections of key-value pairs where the key can be a string, symbol, or other types, and the value has a defined schema.

String Keys

You can define a record with string keys and a specified type for the values.

Example (String Keys with Number Values)

import { Schema } from "effect"

// Define a record schema with string keys and number values
//
//      ┌─── Record$<typeof Schema.String, typeof Schema.Number>
//      ▼
const schema = Schema.Record({ key: Schema.String, value: Schema.Number })

//     ┌─── { readonly [x: string]: number; }
//     ▼
type Type = typeof schema.Type

Symbol Keys

Records can also use symbols as keys.

Example (Symbol Keys with Number Values)

import { Schema } from "effect"

// Define a record schema with symbol keys and number values
const schema = Schema.Record({
  key: Schema.SymbolFromSelf,
  value: Schema.Number
})

//     ┌─── { readonly [x: symbol]: number; }
//     ▼
type Type = typeof schema.Type

Union of Literal Keys

Use a union of literals to restrict keys to a specific set of values.

Example (Union of String Literals as Keys)

import { Schema } from "effect"

// Define a record schema where keys are limited
// to specific string literals ("a" or "b")
const schema = Schema.Record({
  key: Schema.Union(Schema.Literal("a"), Schema.Literal("b")),
  value: Schema.Number
})

//     ┌─── { readonly a: number; readonly b: number; }
//     ▼
type Type = typeof schema.Type

Template Literal Keys

Records can use template literals as keys, allowing for more complex key patterns.

Example (Template Literal Keys with Number Values)

import { Schema } from "effect"

// Define a record schema with keys that match
// the template literal pattern "a${string}"
const schema = Schema.Record({
  key: Schema.TemplateLiteral(Schema.Literal("a"), Schema.String),
  value: Schema.Number
})

//     ┌─── { readonly [x: `a${string}`]: number; }
//     ▼
type Type = typeof schema.Type

Refined Keys

You can refine the key type with additional constraints.

Example (Filtering Keys by Minimum Length)

import { Schema } from "effect"

// Define a record schema where keys are strings with a minimum length of 2
const schema = Schema.Record({
  key: Schema.String.pipe(Schema.minLength(2)),
  value: Schema.Number
})

//     ┌─── { readonly [x: string]: number; }
//     ▼
type Type = typeof schema.Type

Refinements on keys act as filters rather than causing a decoding failure. If a key does not meet the constraints (such as a pattern or minimum length check), it is removed from the decoded output instead of triggering an error.

Example (Keys That Do Not Meet Constraints Are Removed)

import { Schema } from "effect"

const schema = Schema.Record({
  key: Schema.String.pipe(Schema.minLength(2)),
  value: Schema.Number
})

console.log(Schema.decodeUnknownSync(schema)({ a: 1, bb: 2 }))
// Output: { bb: 2 } ("a" is removed because it is too short)

If you want decoding to fail when a key does not meet the constraints, you can set onExcessProperty to "error".

Example (Forcing an Error on Invalid Keys)

import { Schema } from "effect"

const schema = Schema.Record({
  key: Schema.String.pipe(Schema.minLength(2)),
  value: Schema.Number
})

console.log(
  Schema.decodeUnknownSync(schema, { onExcessProperty: "error" })({
    a: 1,
    bb: 2
  })
)
/*
throws:
ParseError: { readonly [x: minLength(2)]: number }
└─ ["a"]
   └─ is unexpected, expected: minLength(2)
*/

Transforming Keys

The Schema.Record API does not support transformations on key schemas. Attempting to apply a transformation to keys will result in an Unsupported key schema error:

Example (Attempting to Transform Keys)

import { Schema } from "effect"

const schema = Schema.Record({
  key: Schema.Trim,
  value: Schema.NumberFromString
})
/*
throws:
Error: Unsupported key schema
schema (Transformation): Trim
*/

Why Key Transformations Are Not Allowed

This restriction exists because transformations can create conflicts if multiple keys map to the same value after transformation. To prevent these issues, key transformations must be handled explicitly by the user.

To modify record keys, you must apply transformations outside of Schema.Record. A common approach is to use Schema.transform to adjust keys during decoding.

Example (Trimming Keys While Decoding)

import { Schema, Record, identity } from "effect"

const schema = Schema.transform(
  // Define the input schema with unprocessed keys
  Schema.Record({
    key: Schema.String,
    value: Schema.NumberFromString
  }),
  // Define the output schema with transformed keys
  Schema.Record({
    key: Schema.Trimmed,
    value: Schema.Number
  }),
  {
    strict: true,
    // Trim keys during decoding
    decode: (record) => Record.mapKeys(record, (key) => key.trim()),
    encode: identity
  }
)

console.log(
  Schema.decodeUnknownSync(schema)({ " key1 ": "1", key2: "2" })
)
// Output: { key1: 1, key2: 2 }

Mutable Records

By default, Schema.Record generates a type marked as readonly. To create a schema for a mutable record, you can use the Schema.mutable function, which makes the record type mutable in a shallow manner.

Example (Creating a Mutable Record Schema)

import { Schema } from "effect"

// Create a schema for a mutable record with string keys and number values
const schema = Schema.mutable(
  Schema.Record({ key: Schema.String, value: Schema.Number })
)

//     ┌─── { [x: string]: number; }
//     ▼
type Type = typeof schema.Type

Exposed Values

You can access the key and value types of a record schema using the key and value properties:

Example (Accessing Key and Value Types)

import { Schema } from "effect"

const schema = Schema.Record({ key: Schema.String, value: Schema.Number })

// Accesses the key
//
//     ┌─── typeof Schema.String
//     ▼
const key = schema.key

// Accesses the value
//
//      ┌─── typeof Schema.Number
//      ▼
const value = schema.value

Structs

Property Signatures

The Schema.Struct constructor defines a schema for an object with specific properties.

Example (Defining a Struct Schema)

This example defines a struct schema for an object with the following properties:

• name: a string
• age: a number

import { Schema } from "effect"

//      ┌─── Schema.Struct<{
//      │      name: typeof Schema.String;
//      │      age: typeof Schema.Number;
//      │    }>
//      ▼
const schema = Schema.Struct({
  name: Schema.String,
  age: Schema.Number
})

// The inferred TypeScript type from the schema
//
//     ┌─── {
//     │      readonly name: string;
//     │      readonly age: number;
//     │    }
//     ▼
type Type = typeof schema.Type

Empty Structs Allow Any Data

Using Schema.Struct({}) results in a TypeScript type {}, which behaves similarly to unknown. This means that any data will be considered valid, as there are no defined constraints.

Index Signatures

The Schema.Struct constructor can optionally accept a list of key/value pairs representing index signatures, allowing you to define additional dynamic properties.

declare const Struct: (props, ...indexSignatures) => Struct<...>

Example (Adding an Index Signature)

import { Schema } from "effect"

// Define a struct with a specific property "a"
// and an index signature allowing additional properties
const schema = Schema.Struct(
  // Defined properties
  { a: Schema.Number },
  // Index signature: allows additional string keys with number values
  { key: Schema.String, value: Schema.Number }
)

// The inferred TypeScript type:
//
//     ┌─── {
//     │      readonly [x: string]: number;
//     │      readonly a: number;
//     │    }
//     ▼
type Type = typeof schema.Type

Example (Using Schema.Record)

You can achieve the same result using Schema.Record:

import { Schema } from "effect"

// Define a struct with a fixed property "a"
// and a dynamic index signature using Schema.Record
const schema = Schema.Struct(
  { a: Schema.Number },
  Schema.Record({ key: Schema.String, value: Schema.Number })
)

// The inferred TypeScript type:
//
//     ┌─── {
//     │      readonly [x: string]: number;
//     │      readonly a: number;
//     │    }
//     ▼
type Type = typeof schema.Type

Multiple Index Signatures

You can define one index signature per key type (string or symbol). Defining multiple index signatures of the same type is not allowed.

Example (Valid Multiple Index Signatures)

import { Schema } from "effect"

// Define a struct with a fixed property "a"
// and valid index signatures for both strings and symbols
const schema = Schema.Struct(
  { a: Schema.Number },
  // String index signature
  { key: Schema.String, value: Schema.Number },
  // Symbol index signature
  { key: Schema.SymbolFromSelf, value: Schema.Number }
)

// The inferred TypeScript type:
//
//     ┌─── {
//     │      readonly [x: string]: number;
//     │      readonly [x: symbol]: number;
//     │      readonly a: number;
//     │    }
//     ▼
type Type = typeof schema.Type

Defining multiple index signatures of the same key type (string or symbol) will cause an error.

Example (Invalid Multiple Index Signatures)

import { Schema } from "effect"

Schema.Struct(
  { a: Schema.Number },
  // Attempting to define multiple string index signatures
  { key: Schema.String, value: Schema.Number },
  { key: Schema.String, value: Schema.Boolean }
)
/*
throws:
Error: Duplicate index signature
details: string index signature
*/

Conflicting Index Signatures

When defining schemas with index signatures, conflicts can arise if a fixed property has a different type than the values allowed by the index signature. This can lead to unexpected TypeScript behavior.

Example (Conflicting Index Signature)

import { Schema } from "effect"

// Attempting to define a struct with a conflicting index signature
// - The fixed property "a" is a number
// - The index signature requires all values to be strings
const schema = Schema.Struct(
  { a: Schema.String },
  { key: Schema.String, value: Schema.Number }
)

// ❌ Incorrect TypeScript type:
//
//     ┌─── {
//     │      readonly [x: string]: number;
//     │      readonly a: string;
//     │    }
//     ▼
type Type = typeof schema.Type

The TypeScript compiler flags this as an error when defining the type manually:

// This type is invalid because the index signature
// conflicts with the fixed property `a`
type Test = {
  readonly a: string
Error ts(2411)  ― Property 'a' of type 'string' is not assignable to 'string' index type 'number'.
  readonly [x: string]: number
}

This happens because TypeScript does not allow an index signature to contradict a fixed property.

Workaround for Conflicting Index Signatures

When working with schemas, a conflict can occur if a fixed property has a different type than the values allowed by an index signature. This situation often arises when dealing with external APIs that do not follow strict TypeScript conventions.

To prevent conflicts, you can separate the fixed properties from the indexed properties and handle them as distinct parts of the schema.

Example (Extracting Fixed and Indexed Properties)

Consider an object where:

• "a" is a fixed property of type string.
• All other keys store numbers, which conflict with "a".

// This type is invalid because the index signature
// conflicts with the fixed property `a`
type Test = {
  a: string
Error ts(2411)  ― Property 'a' of type 'string' is not assignable to 'string' index type 'number'.
  [x: string]: number
}

To avoid this issue, we can separate the properties into two distinct types:

// Fixed properties schema
type FixedProperties = {
  readonly a: string
}

// Index signature properties schema
type IndexSignatureProperties = {
  readonly [x: string]: number
}

// The final output groups both properties in a tuple
type OutputData = readonly [FixedProperties, IndexSignatureProperties]

By using Schema.transform and Schema.compose, you can preprocess the input data before validation. This approach ensures that fixed properties and index signature properties are treated independently.

import { Schema } from "effect"

// Define a schema for the fixed property "a"
const FixedProperties = Schema.Struct({
  a: Schema.String
})

// Define a schema for index signature properties
const IndexSignatureProperties = Schema.Record({
  // Exclude keys that are already present in FixedProperties
  key: Schema.String.pipe(
    Schema.filter(
      (key) => !Object.keys(FixedProperties.fields).includes(key)
    )
  ),
  value: Schema.Number
})

// Create a schema that duplicates an object into two parts
const Duplicate = Schema.transform(
  Schema.Object,
  Schema.Tuple(Schema.Object, Schema.Object),
  {
    strict: true,
    // Create a tuple containing the input twice
    decode: (a) => [a, a] as const,
    // Merge both parts back when encoding
    encode: ([a, b]) => ({ ...a, ...b })
  }
)

//      ┌─── Schema<readonly [
//      |      { readonly a: string; },
//      |      { readonly [x: string]: number; }
//      |    ], object>
//      ▼
const Result = Schema.compose(
  Duplicate,
  Schema.Tuple(FixedProperties, IndexSignatureProperties).annotations({
    parseOptions: { onExcessProperty: "ignore" }
  })
)

// Decoding: Separates fixed and indexed properties
console.log(Schema.decodeUnknownSync(Result)({ a: "a", b: 1, c: 2 }))
// Output: [ { a: 'a' }, { b: 1, c: 2 } ]

// Encoding: Combines them back into an object
console.log(Schema.encodeSync(Result)([{ a: "a" }, { b: 1, c: 2 }]))
// Output: { a: 'a', b: 1, c: 2 }

Exposed Values

You can access the fields and records of a struct schema using the fields and records properties:

Example (Accessing Fields and Records)

import { Schema } from "effect"

const schema = Schema.Struct(
  { a: Schema.Number },
  Schema.Record({ key: Schema.String, value: Schema.Number })
)

// Accesses the fields
//
//      ┌─── { readonly a: typeof Schema.Number; }
//      ▼
const fields = schema.fields

// Accesses the records
//
//      ┌─── readonly [Schema.Record$<typeof Schema.String, typeof Schema.Number>]
//      ▼
const records = schema.records

Mutable Structs

By default, Schema.Struct generates a type with properties marked as readonly. To create a mutable version of the struct, use the Schema.mutable function, which makes the properties mutable in a shallow manner.

Example (Creating a Mutable Struct Schema)

import { Schema } from "effect"

const schema = Schema.mutable(
  Schema.Struct({ a: Schema.String, b: Schema.Number })
)

//     ┌─── { a: string; b: number; }
//     ▼
type Type = typeof schema.Type

Tagged Structs

In TypeScript tags help to enhance type discrimination and pattern matching by providing a simple yet powerful way to define and recognize different data types.

What is a Tag?

A tag is a literal value added to data structures, commonly used in structs, to distinguish between various object types or variants within tagged unions. This literal acts as a discriminator, making it easier to handle and process different types of data correctly and efficiently.

Using the tag Constructor

The Schema.tag constructor is specifically designed to create a property signature that holds a specific literal value, serving as the discriminator for object types.

Example (Defining a Tagged Struct)

import { Schema } from "effect"

const User = Schema.Struct({
  _tag: Schema.tag("User"),
  name: Schema.String,
  age: Schema.Number
})

//     ┌─── { readonly _tag: "User"; readonly name: string; readonly age: number; }
//     ▼
type Type = typeof User.Type

console.log(User.make({ name: "John", age: 44 }))
/*
Output:
{ _tag: 'User', name: 'John', age: 44 }
*/

In the example above, Schema.tag("User") attaches a _tag property to the User struct schema, effectively labeling objects of this struct type as “User”. This label is automatically applied when using the make method to create new instances, simplifying object creation and ensuring consistent tagging.

Simplifying Tagged Structs with TaggedStruct

The Schema.TaggedStruct constructor streamlines the process of creating tagged structs by directly integrating the tag into the struct definition. This method provides a clearer and more declarative approach to building data structures with embedded discriminators.

Example (Using TaggedStruct for a Simplified Tagged Struct)

import { Schema } from "effect"

const User = Schema.TaggedStruct("User", {
  name: Schema.String,
  age: Schema.Number
})

// `_tag` is automatically applied when constructing an instance
console.log(User.make({ name: "John", age: 44 }))
// Output: { _tag: 'User', name: 'John', age: 44 }

// `_tag` is required when decoding from an unknown source
console.log(Schema.decodeUnknownSync(User)({ name: "John", age: 44 }))
/*
throws:
ParseError: { readonly _tag: "User"; readonly name: string; readonly age: number }
└─ ["_tag"]
   └─ is missing
*/

In this example:

• The _tag property is optional when constructing an instance with make, allowing the schema to automatically apply it.
• When decoding unknown data, _tag is required to ensure correct type identification. This distinction between instance construction and decoding is useful for preserving the tag’s role as a type discriminator while simplifying instance creation.

If you need _tag to be applied automatically during decoding as well, you can create a customized version of Schema.TaggedStruct:

Example (Custom TaggedStruct with _tag Applied during Decoding)

import type { SchemaAST } from "effect"
import { Schema } from "effect"

const TaggedStruct = <
  Tag extends SchemaAST.LiteralValue,
  Fields extends Schema.Struct.Fields
>(
  tag: Tag,
  fields: Fields
) =>
  Schema.Struct({
    _tag: Schema.Literal(tag).pipe(
      Schema.optional,
      Schema.withDefaults({
        constructor: () => tag, // Apply _tag during instance construction
        decoding: () => tag // Apply _tag during decoding
      })
    ),
    ...fields
  })

const User = TaggedStruct("User", {
  name: Schema.String,
  age: Schema.Number
})

console.log(User.make({ name: "John", age: 44 }))
// Output: { _tag: 'User', name: 'John', age: 44 }

console.log(Schema.decodeUnknownSync(User)({ name: "John", age: 44 }))
// Output: { _tag: 'User', name: 'John', age: 44 }

Multiple Tags

While a primary tag is often sufficient, TypeScript allows you to define multiple tags for more complex data structuring needs. Here’s an example demonstrating the use of multiple tags within a single struct:

Example (Adding Multiple Tags to a Struct)

This example defines a product schema with a primary tag ("Product") and an additional category tag ("Electronics"), adding further specificity to the data structure.

import { Schema } from "effect"

const Product = Schema.TaggedStruct("Product", {
  category: Schema.tag("Electronics"),
  name: Schema.String,
  price: Schema.Number
})

// `_tag` and `category` are optional when creating an instance
console.log(Product.make({ name: "Smartphone", price: 999 }))
/*
Output:
{
  _tag: 'Product',
  category: 'Electronics',
  name: 'Smartphone',
  price: 999
}
*/

instanceOf

When you need to define a schema for your custom data type defined through a class, the most convenient and fast way is to use the Schema.instanceOf constructor.

Example (Defining a Schema with instanceOf)

import { Schema } from "effect"

// Define a custom class
class MyData {
  constructor(readonly name: string) {}
}

// Create a schema for the class
const MyDataSchema = Schema.instanceOf(MyData)

//     ┌─── MyData
//     ▼
type Type = typeof MyDataSchema.Type

console.log(Schema.decodeUnknownSync(MyDataSchema)(new MyData("name")))
// Output: MyData { name: 'name' }

console.log(Schema.decodeUnknownSync(MyDataSchema)({ name: "name" }))
/*
throws:
ParseError: Expected MyData, actual {"name":"name"}
*/

The Schema.instanceOf constructor is just a lightweight wrapper of the Schema.declare API, which is the primitive in effect/Schema for declaring new custom data types.

Private Constructors

Note that Schema.instanceOf can only be used for classes that expose a public constructor. If you try to use it with classes that, for some reason, have marked the constructor as private, you’ll receive a TypeScript error:

Example (Error With Private Constructors)

import { Schema } from "effect"

class MyData {
  static make = (name: string) => new MyData(name)
  private constructor(readonly name: string) {}
}

// @ts-expect-error
const MyDataSchema = Schema.instanceOf(MyData)
/*
Argument of type 'typeof MyData' is not assignable to parameter of type 'abstract new (...args: any) => any'.
  Cannot assign a 'private' constructor type to a 'public' constructor type.ts(2345)
*/

In such cases, you cannot use Schema.instanceOf, and you must rely on Schema.declare like this:

Example (Using Schema.declare With Private Constructors)

import { Schema } from "effect"

class MyData {
  static make = (name: string) => new MyData(name)
  private constructor(readonly name: string) {}
}

const MyDataSchema = Schema.declare(
  (input: unknown): input is MyData => input instanceof MyData
).annotations({ identifier: "MyData" })

console.log(Schema.decodeUnknownSync(MyDataSchema)(MyData.make("name")))
// Output: MyData { name: 'name' }

console.log(Schema.decodeUnknownSync(MyDataSchema)({ name: "name" }))
/*
throws:
ParseError: Expected MyData, actual {"name":"name"}
*/

Validating Fields of the Instance

To validate the fields of a class instance, you can use a filter. This approach combines instance validation with additional checks on the instance’s fields.

Example (Adding Field Validation to an Instance Schema)

import { Either, ParseResult, Schema } from "effect"

class MyData {
  constructor(readonly name: string) {}
}

const MyDataFields = Schema.Struct({
  name: Schema.NonEmptyString
})

// Define a schema for the class instance with additional field validation
const MyDataSchema = Schema.instanceOf(MyData).pipe(
  Schema.filter((a, options) =>
    // Validate the fields of the instance
    ParseResult.validateEither(MyDataFields)(a, options).pipe(
      // Invert success and failure for filtering
      Either.flip,
      // Return undefined if validation succeeds, or an error if it fails
      Either.getOrUndefined
    )
  )
)

// Example: Valid instance
console.log(Schema.validateSync(MyDataSchema)(new MyData("John")))
// Output: MyData { name: 'John' }

// Example: Invalid instance (empty name)
console.log(Schema.validateSync(MyDataSchema)(new MyData("")))
/*
throws:
ParseError: { MyData | filter }
└─ Predicate refinement failure
   └─ { readonly name: NonEmptyString }
      └─ ["name"]
         └─ NonEmptyString
            └─ Predicate refinement failure
               └─ Expected a non empty string, actual ""
*/

Picking

The pick static function available on each struct schema can be used to create a new Struct by selecting specific properties from an existing Struct.

Example (Picking Properties from a Struct)

import { Schema } from "effect"

// Define a struct schema with properties "a", "b", and "c"
const MyStruct = Schema.Struct({
  a: Schema.String,
  b: Schema.Number,
  c: Schema.Boolean
})

// Create a new schema that picks properties "a" and "c"
//
//      ┌─── Struct<{
//      |      a: typeof Schema.String;
//      |      c: typeof Schema.Boolean;
//      |    }>
//      ▼
const PickedSchema = MyStruct.pick("a", "c")

The Schema.pick function can be applied more broadly beyond just Struct types, such as with unions of schemas. However it returns a generic SchemaClass.

Example (Picking Properties from a Union)

import { Schema } from "effect"

// Define a union of two struct schemas
const MyUnion = Schema.Union(
  Schema.Struct({ a: Schema.String, b: Schema.String, c: Schema.String }),
  Schema.Struct({ a: Schema.Number, b: Schema.Number, d: Schema.Number })
)

// Create a new schema that picks properties "a" and "b"
//
//      ┌─── SchemaClass<{
//      |      readonly a: string | number;
//      |      readonly b: string | number;
//      |    }>
//      ▼
const PickedSchema = MyUnion.pipe(Schema.pick("a", "b"))

Omitting

The omit static function available in each struct schema can be used to create a new Struct by excluding particular properties from an existing Struct.

Example (Omitting Properties from a Struct)

import { Schema } from "effect"

// Define a struct schema with properties "a", "b", and "c"
const MyStruct = Schema.Struct({
  a: Schema.String,
  b: Schema.Number,
  c: Schema.Boolean
})

// Create a new schema that omits property "b"
//
//      ┌─── Schema.Struct<{
//      |      a: typeof Schema.String;
//      |      c: typeof Schema.Boolean;
//      |    }>
//      ▼
const PickedSchema = MyStruct.omit("b")

The Schema.omit function can be applied more broadly beyond just Struct types, such as with unions of schemas. However it returns a generic Schema.

Example (Omitting Properties from a Union)

import { Schema } from "effect"

// Define a union of two struct schemas
const MyUnion = Schema.Union(
  Schema.Struct({ a: Schema.String, b: Schema.String, c: Schema.String }),
  Schema.Struct({ a: Schema.Number, b: Schema.Number, d: Schema.Number })
)

// Create a new schema that omits property "b"
//
//      ┌─── SchemaClass<{
//      |      readonly a: string | number;
//      |    }>
//      ▼
const PickedSchema = MyUnion.pipe(Schema.omit("b"))

partial

The Schema.partial function makes all properties within a schema optional.

Example (Making All Properties Optional)

import { Schema } from "effect"

// Create a schema with an optional property "a"
const schema = Schema.partial(Schema.Struct({ a: Schema.String }))

//     ┌─── { readonly a?: string | undefined; }
//     ▼
type Type = typeof schema.Type

By default, the Schema.partial operation adds undefined to the type of each property. If you want to avoid this, you can use Schema.partialWith and pass { exact: true } as an argument.

Example (Defining an Exact Partial Schema)

import { Schema } from "effect"

// Create a schema with an optional property "a" without allowing undefined
const schema = Schema.partialWith(
  Schema.Struct({
    a: Schema.String
  }),
  { exact: true }
)

//     ┌─── { readonly a?: string; }
//     ▼
type Type = typeof schema.Type

required

The Schema.required function ensures that all properties in a schema are mandatory.

Example (Making All Properties Required)

import { Schema } from "effect"

// Create a schema and make all properties required
const schema = Schema.required(
  Schema.Struct({
    a: Schema.optionalWith(Schema.String, { exact: true }),
    b: Schema.optionalWith(Schema.Number, { exact: true })
  })
)

//     ┌─── { readonly a: string; readonly b: number; }
//     ▼
type Type = typeof schema.Type

In this example, both a and b are made required, even though they were initially defined as optional.

keyof

The Schema.keyof operation creates a schema that represents the keys of a given object schema.

Example (Extracting Keys from an Object Schema)

import { Schema } from "effect"

const schema = Schema.Struct({
  a: Schema.String,
  b: Schema.Number
})

const keys = Schema.keyof(schema)

//     ┌─── "a" | "b"
//     ▼
type Type = typeof keys.Type