@satankebab/ts-pattern

TS-Pattern

The exhaustive Pattern Matching library for TypeScript with smart type inference.

import { match, P } from 'ts-pattern';

type Data =
  | { type: 'text'; content: string }
  | { type: 'img'; src: string };

type Result =
  | { type: 'ok'; data: Data }
  | { type: 'error'; error: Error };

const result: Result = ...;

const html = match(result)
  .with({ type: 'error' }, () => <p>Oups! An error occured</p>)
  .with({ type: 'ok', data: { type: 'text' } }, (res) => <p>{res.data.content}</p>)
  .with({ type: 'ok', data: { type: 'img', src: P.select() } }, (src) => <img src={src} />)
  .exhaustive();

About

Write better and safer conditions. Pattern matching lets you express complex conditions in a single, compact expression. Your code becomes shorter and more readable. Exhaustiveness checking ensures you haven’t forgotten any possible case.

Animation by @nicoespeon

Features

• Pattern-match on any data structure: nested Objects, Arrays, Tuples, Sets, Maps and all primitive types.
• Typesafe, with helpful type inference.
• Exhaustiveness checking support, enforcing that you are matching every possible case with .exhaustive().
• Use patterns to validate the shape of your data with isMatching.
• Expressive API, with catch-all and type specific wildcards: P._, P.string, P.number, etc.
• Supports predicates, unions, intersections and exclusion patterns for non-trivial cases.
• Supports properties selection, via the P.select(name?) function.
• Tiny bundle footprint (only ~2kB).

What is Pattern Matching?

Pattern Matching is a code-branching technique coming from functional programming languages that's more powerful and often less verbose than imperative alternatives (if/else/switch statements), especially for complex conditions.

Pattern Matching is implemented in Haskell, Rust, Swift, Elixir, and many other languages. There is a tc39 proposal to add Pattern Matching to EcmaScript, but it is still in stage 1 and isn't likely to land before several years. Luckily, pattern matching can be implemented in userland. ts-pattern Provides a typesafe pattern matching implementation that you can start using today.

Read the introduction blog post: Bringing Pattern Matching to TypeScript 🎨 Introducing TS-Pattern

Installation

Via npm

npm install ts-pattern

Via yarn

yarn add ts-pattern

Via pnpm

pnpm add ts-pattern

Via Bun

bun add ts-pattern

Compatibility with different TypeScript versions

TS-Pattern assumes that Strict Mode is enabled in your tsconfig.json file.

ts-pattern	TypeScript v5+	TypeScript v4.5+	TypeScript v4.2+
v5.x (Docs) (Migration Guide)	✅	❌	❌
v4.x (Docs) (Migration Guide)	✅	✅	❌
v3.x (Docs)	✅	✅	✅

• ✅ Full support
• ❌ Not supported

Documentation

• Code Sandbox Examples
• Getting Started
• API Reference
  • match
  • .with
  • .when
  • .returnType
  • .exhaustive
  • .otherwise
  • .run
  • isMatching
  • Patterns
    • Literals
    • Wildcards
    • Objects
    • Tuples (arrays)
    • Sets
    • Maps
    • P.array patterns
    • P.when patterns
    • P.not patterns
    • P.select patterns
    • P.optional patterns
    • P.instanceOf patterns
    • P.union patterns
    • P.intersection patterns
    • P.string predicates
    • P.number and P.bigint predicates
  • Types
    • P.infer
    • P.Pattern
    • Type inference
• Inspirations

Code Sandbox Examples

• Basic Demo
• React gif fetcher app Demo
• React.useReducer Demo
• Handling untyped API response Demo
• P.when Guard Demo
• P.not Pattern Demo
• P.select Pattern Demo
• P.union Pattern Demo

Getting Started

As an example, let's create a state reducer for a frontend application that fetches some data.

Example: a state reducer with ts-pattern

Our application can be in four different states: idle, loading, success and error. Depending on which state we are in, some events can occur. Here are all the possible types of event our application can respond to: fetch, success, error and cancel.

I use the word event but you can replace it with action if you are used to Redux's terminology.

type State =
  | { status: 'idle' }
  | { status: 'loading'; startTime: number }
  | { status: 'success'; data: string }
  | { status: 'error'; error: Error };

type Event =
  | { type: 'fetch' }
  | { type: 'success'; data: string }
  | { type: 'error'; error: Error }
  | { type: 'cancel' };

Even though our application can handle 4 events, only a subset of these events make sense for each given state. For instance we can only cancel a request if we are currently in the loading state. To avoid unwanted state changes that could lead to bugs, we want our state reducer function to branch on both the state and the event, and return a new state.

This is a case where match really shines. Instead of writing nested switch statements, we can use pattern matching to simultaneously check the state and the event object:

import { match, P } from 'ts-pattern';

const reducer = (state: State, event: Event) =>
  match([state, event])
    .returnType<State>()
    .with(
      [{ status: 'loading' }, { type: 'success' }],
      ([_, event]) => ({ status: 'success', data: event.data })
    )
    .with(
      [{ status: 'loading' }, { type: 'error', error: P.select() }],
      (error) => ({ status: 'error', error })
    )
    .with(
      [{ status: P.not('loading') }, { type: 'fetch' }],
      () => ({ status: 'loading', startTime: Date.now() })
    )
    .with(
      [
        {
          status: 'loading',
          startTime: P.when((t) => t + 2000 < Date.now()),
        },
        { type: 'cancel' },
      ],
      () => ({ status: 'idle' })
    )
    .with(P._, () => state)
    .exhaustive();

There's a lot going on, so let's go through this code bit by bit:

match(value)

match takes a value and returns a builder on which you can add your pattern matching cases.

match([state, event])

It's also possible to specify the input and output type explicitly with match<Input, Output>(...), but this is usually unnecessary, as TS-Pattern is able to infer them.

.returnType<OutputType>()

.returnType is an optional method that you can call if you want to force all following code-branches to return a value of a specific type. It takes a single type parameter, provided between <AngleBrackets>.

  .returnType<State>()

Here, we use this method to make sure all branches return a valid State object.

.with(pattern, handler)

Then we add a first with clause:

  .with(
    [{ status: 'loading' }, { type: 'success' }],
    ([state, event]) => ({
      // `state` is inferred as { status: 'loading' }
      // `event` is inferred as { type: 'success', data: string }
      status: 'success',
      data: event.data,
    })
  )

The first argument is the pattern: the shape of value you expect for this branch.

The second argument is the handler function: the code branch that will be called if the input value matches the pattern.

The handler function takes the input value as first parameter with its type narrowed down to what the pattern matches.

P.select(name?)

In the second with clause, we use the P.select function:

  .with(
    [
      { status: 'loading' },
      { type: 'error', error: P.select() }
    ],
    (error) => ({ status: 'error', error })
  )

P.select() lets you extract a piece of your input value and inject it into your handler. It is pretty useful when pattern matching on deep data structures because it avoids the hassle of destructuring your input in your handler.

Since we didn't pass any name to P.select(), It will inject the event.error property as first argument to the handler function. Note that you can still access the full input value with its type narrowed by your pattern as second argument of the handler function:

  .with(
    [
      { status: 'loading' },
      { type: 'error', error: P.select() }
    ],
    (error, stateAndEvent) => {
      // error: Error
      // stateAndEvent: [{ status: 'loading' }, { type: 'error', error: Error }]
    }
  )

In a pattern, we can only have a single anonymous selection. If you need to select more properties on your input data structure, you will need to give them names:

.with(
    [
      { status: 'success', data: P.select('prevData') },
      { type: 'error', error: P.select('err') }
    ],
    ({ prevData, err }) => {
      // Do something with (prevData: string) and (err: Error).
    }
  )

Each named selection will be injected inside a selections object, passed as first argument to the handler function. Names can be any strings.

P.not(pattern)

If you need to match on everything but a specific value, you can use a P.not(<pattern>) pattern. it's a function taking a pattern and returning its opposite:

  .with(
    [{ status: P.not('loading') }, { type: 'fetch' }],
    () => ({ status: 'loading' })
  )

P.when() and guard functions

Sometimes, we need to make sure our input value respects a condition that can't be expressed by a pattern. For example, imagine you need to check that a number is positive. In these cases, we can use guard functions: functions taking a value and returning a boolean.

With TS-Pattern, there are two ways to use a guard function:

• use P.when(<guard function>) inside one of your patterns
• pass it as second parameter to .with(...)

using P.when(predicate)

  .with(
    [
      {
        status: 'loading',
        startTime: P.when((t) => t + 2000 < Date.now()),
      },
      { type: 'cancel' },
    ],
    () => ({ status: 'idle' })
  )

Passing a guard function to .with(...)

.with optionally accepts a guard function as second parameter, between the pattern and the handler callback:

  .with(
    [{ status: 'loading' }, { type: 'cancel' }],
    ([state, event]) => state.startTime + 2000 < Date.now(),
    () => ({ status: 'idle' })
  )

This pattern will only match if the guard function returns true.

the P._ wildcard

P._ will match any value. You can use it either at the top level, or within another pattern.

  .with(P._, () => state)

  // You could also use it inside another pattern:
  .with([P._, P._], () => state)

  // at any level:
  .with([P._, { type: P._ }], () => state)

.exhaustive(), .otherwise() and .run()

  .exhaustive();

.exhaustive() executes the pattern matching expression, and returns the result. It also enables exhaustiveness checking, making sure we don't forget any possible case in our input value. This extra type safety is very nice because forgetting a case is an easy mistake to make, especially in an evolving code-base.

Note that exhaustive pattern matching is optional. It comes with the trade-off of having slightly longer compilation times because the type checker has more work to do.

Alternatively, you can use .otherwise(), which takes an handler function returning a default value. .otherwise(handler) is equivalent to .with(P._, handler).exhaustive().

  .otherwise(() => state);

Matching several patterns

As you may know, switch statements allow handling several cases with the same code block:

switch (type) {
  case 'text':
  case 'span':
  case 'p':
    return 'text';

  case 'btn':
  case 'button':
    return 'button';
}

Similarly, ts-pattern lets you pass several patterns to .with() and if one of these patterns matches your input, the handler function will be called:

const sanitize = (name: string) =>
  match(name)
    .with('text', 'span', 'p', () => 'text')
    .with('btn', 'button', () => 'button')
    .otherwise(() => name);

sanitize('span'); // 'text'
sanitize('p'); // 'text'
sanitize('button'); // 'button'

As you might expect, this also works with more complex patterns than strings and exhaustiveness checking works as well.

API Reference

match

match(value);

Create a Match object on which you can later call .with, .when, .otherwise and .run.

Signature

function match<TInput, TOutput>(input: TInput): Match<TInput, TOutput>;

Arguments

• input
  • Required
  • the input value your patterns will be tested against.

.with

match(...)
  .with(pattern, [...patterns], handler)

Signature

function with(
  pattern: Pattern<TInput>,
  handler: (selections: Selections<TInput>, value: TInput) => TOutput
): Match<TInput, TOutput>;

// Overload for multiple patterns
function with(
  pattern1: Pattern<TInput>,
  ...patterns: Pattern<TInput>[],
  // no selection object is provided when using multiple patterns
  handler: (value: TInput) => TOutput
): Match<TInput, TOutput>;

// Overload for guard functions
function with(
  pattern: Pattern<TInput>,
  when: (value: TInput) => unknown,
  handler: (
    selection: Selection<TInput>,
    value: TInput
  ) => TOutput
): Match<TInput, TOutput>;

Arguments

• pattern: Pattern<TInput>
  • Required
  • The pattern your input must match for the handler to be called.
  • See all valid patterns below
  • If you provide several patterns before providing the handler, the with clause will match if one of the patterns matches.
• when: (value: TInput) => unknown
  • Optional
  • Additional condition the input must satisfy for the handler to be called.
  • The input will match if your guard function returns a truthy value.
  • TInput might be narrowed to a more precise type using the pattern.
• handler: (selections: Selections<TInput>, value: TInput) => TOutput
  • Required
  • Function called when the match conditions are satisfied.
  • All handlers on a single match case must return values of the same type, TOutput.
  • selections is an object of properties selected from the input with the select function.
  • TInput might be narrowed to a more precise type using the pattern.

.when

match(...)
  .when(predicate, handler)

Signature

function when(
  predicate: (value: TInput) => unknown,
  handler: (value: TInput) => TOutput
): Match<TInput, TOutput>;

Arguments

• predicate: (value: TInput) => unknown
  • Required
  • Condition the input must satisfy for the handler to be called.
• handler: (value: TInput) => TOutput
  • Required
  • Function called when the predicate condition is satisfied.
  • All handlers on a single match case must return values of the same type, TOutput.

.returnType

match(...)
  .returnType<string>()
  .with(..., () => "has to be a string")
  .with(..., () => "Oops".length)
  //               ~~~~~~~~~~~~~ ❌ `number` isn't a string!

The .returnType<SomeType>() method allows you to control the return type of all of your branches of code. It accepts a single type parameter that will be used as the return type of your match expression. All code branches must return values assignable to this type.

Signature

function returnType<TOutputOverride>(): Match<TInput, TOutputOverride>;

Type arguments

• TOutputOverride
  • The type that your match expression will return. All branches must return values assignable to it.

.exhaustive

match(...)
  .with(...)
  .exhaustive()

Runs the pattern-matching expression and returns its result. It also enables exhaustiveness checking, making sure at compile time that we have handled all possible cases.

Signature

function exhaustive(): TOutput;

Example

type Permission = 'editor' | 'viewer';
type Plan = 'basic' | 'pro';

const fn = (org: Plan, user: Permission) =>
  match([org, user])
    .with(['basic', 'viewer'], () => {})
    .with(['basic', 'editor'], () => {})
    .with(['pro', 'viewer'], () => {})
    // Fails with `NonExhaustiveError<['pro', 'editor']>`
    // because the `['pro', 'editor']` case isn't handled.
    .exhaustive();

const fn2 = (org: Plan, user: Permission) =>
  match([org, user])
    .with(['basic', 'viewer'], () => {})
    .with(['basic', 'editor'], () => {})
    .with(['pro', 'viewer'], () => {})
    .with(['pro', 'editor'], () => {})
    .exhaustive(); // Works!

.otherwise

match(...)
  .with(...)
  .otherwise(defaultHandler)

Runs the pattern-matching expression with a default handler which will be called if no previous .with() clause match the input value, and returns the result.

Signature

function otherwise(defaultHandler: (value: TInput) => TOutput): TOutput;

Arguments

• defaultHandler: (value: TInput) => TOutput
  • Required
  • Function called if no pattern matched the input value.
  • Think of it as the default: case of switch statements.
  • All handlers on a single match case must return values of the same type, TOutput.

.run

match(...)
  .with(...)
  .run()

returns the result of the pattern-matching expression, or throws if no pattern matched the input. .run() is similar to .exhaustive(), but is unsafe because exhaustiveness is not checked at compile time, so you have no guarantees that all cases are indeed covered. Use at your own risks.

Signature

function run(): TOutput;

isMatching

if (isMatching(pattern, value))  {
  ...
}

isMatching is a type guard function which checks if a pattern matches a given value. It is curried, which means it can be used in two ways.

With a single argument:

import { isMatching, P } from 'ts-pattern';

const isBlogPost = isMatching({
  type: 'blogpost',
  title: P.string,
  description: P.string,
});

if (isBlogPost(value)) {
  // value: { type: 'blogpost', title: string, description: string }
}

With two arguments:

const blogPostPattern = {
  type: 'blogpost',
  title: P.string,
  description: P.string,
} as const;

if (isMatching(blogPostPattern, value)) {
  // value: { type: 'blogpost', title: string, description: string }
}

Signature

export function isMatching<p extends Pattern<any>>(
  pattern: p
): (value: any) => value is InvertPattern<p>;
export function isMatching<p extends Pattern<any>>(
  pattern: p,
  value: any
): value is InvertPattern<p>;

Arguments

• pattern: Pattern<any>
  • Required
  • The pattern a value should match.
• value?: any
  • Optional
  • if a value is given as second argument, isMatching will return a boolean telling us whether the pattern matches the value or not.
  • if we only give the pattern to the function, isMatching will return another type guard function taking a value and returning a boolean which tells us whether the pattern matches the value or not.

Patterns

A pattern is a description of the expected shape of your input value.

Patterns can be regular JavaScript values ("some string", 10, true, ...), data structures (objects, arrays, ...), wildcards (P._, P.string, P.number, ...), or special matcher functions (P.not, P.when, P.select, ...).

All wildcards and matcher functions can be imported either as Pattern or as P from the ts-pattern module.

import { match, Pattern } from 'ts-pattern';

const toString = (value: unknown): string =>
  match(value)
    .with(Pattern.string, (str) => str)
    .with(Pattern.number, (num) => num.toFixed(2))
    .with(Pattern.boolean, (bool) => `${bool}`)
    .otherwise(() => 'Unknown');

Or

import { match, P } from 'ts-pattern';

const toString = (value: unknown): string =>
  match(value)
    .with(P.string, (str) => str)
    .with(P.number, (num) => num.toFixed(2))
    .with(P.boolean, (bool) => `${bool}`)
    .otherwise(() => 'Unknown');

If your input isn't typed, (if it's a any or a unknown), you are free to use any possible pattern. Your handler will infer the input type from the shape of your pattern.

Literals

Literals are primitive JavaScript values, like numbers, strings, booleans, bigints, symbols, null, undefined, or NaN.

import { match } from 'ts-pattern';

const input: unknown = 2;

const output = match(input)
  .with(2, () => 'number: two')
  .with(true, () => 'boolean: true')
  .with('hello', () => 'string: hello')
  .with(undefined, () => 'undefined')
  .with(null, () => 'null')
  .with(NaN, () => 'number: NaN')
  .with(20n, () => 'bigint: 20n')
  .otherwise(() => 'something else');

console.log(output);
// => 'number: two'

Objects

Patterns can be objects containing sub-patterns. An object pattern will match If and only if the input value is an object, contains all properties the pattern defines and each property matches the corresponding sub-pattern.

import { match } from 'ts-pattern';

type Input =
  | { type: 'user'; name: string }
  | { type: 'image'; src: string }
  | { type: 'video'; seconds: number };

let input: Input = { type: 'user', name: 'Gabriel' };

const output = match(input)
  .with({ type: 'image' }, () => 'image')
  .with({ type: 'video', seconds: 10 }, () => 'video of 10 seconds.')
  .with({ type: 'user' }, ({ name }) => `user of name: ${name}`)
  .otherwise(() => 'something else');

console.log(output);
// => 'user of name: Gabriel'

Tuples (arrays)

In TypeScript, Tuples are arrays with a fixed number of elements that can be of different types. You can pattern-match on tuples using a tuple pattern. A tuple pattern will match if the input value is an array of the same length, and each item matches the corresponding sub-pattern.

import { match, P } from 'ts-pattern';

type Input =
  | [number, '+', number]
  | [number, '-', number]
  | [number, '*', number]
  | ['-', number];

const input: Input = [3, '*', 4];

const output = match(input)
  .with([P._, '+', P._], ([x, , y]) => x + y)
  .with([P._, '-', P._], ([x, , y]) => x - y)
  .with([P._, '*', P._], ([x, , y]) => x * y)
  .with(['-', P._], ([, x]) => -x)
  .otherwise(() => NaN);

console.log(output);
// => 12

Wildcards

P._ wildcard

The P._ pattern will match any value. You can also use P.any, which is an alias to P._.

import { match, P } from 'ts-pattern';

const input = 'hello';

const output = match(input)
  .with(P._, () => 'It will always match')
  // OR
  .with(P.any, () => 'It will always match')
  .otherwise(() => 'This string will never be used');

console.log(output);
// => 'It will always match'

P.string wildcard

The P.string pattern will match any value of type string.

import { match, P } from 'ts-pattern';

const input = 'hello';

const output = match(input)
  .with('bonjour', () => 'Won‘t match')
  .with(P.string, () => 'it is a string!')
  .exhaustive();

console.log(output);
// => 'it is a string!'

P.number wildcard

The P.number pattern will match any value of type number.

import { match, P } from 'ts-pattern';

const input = 2;

const output = match<number | string>(input)
  .with(P.string, () => 'it is a string!')
  .with(P.number, () => 'it is a number!')
  .exhaustive();

console.log(output);
// => 'it is a number!'

P.boolean wildcard

The P.boolean pattern will match any value of type boolean.

import { match, P } from 'ts-pattern';

const input = true;

const output = match<number | string | boolean>(input)
  .with(P.string, () => 'it is a string!')
  .with(P.number, () => 'it is a number!')
  .with(P.boolean, () => 'it is a boolean!')
  .exhaustive();

console.log(output);
// => 'it is a boolean!'

P.nullish wildcard

The P.nullish pattern will match any value of type null or undefined.

Even though null and undefined can be used as literal patterns, sometimes they appear in a union together (e.g. null | undefined | string) and you may want to treat them as equivalent using P.nullish.

import { match, P } from 'ts-pattern';

const input = null;

const output = match<number | null | undefined>(input)
  .with(P.number, () => 'it is a number!')
  .with(P.nullish, () => 'it is either null or undefined!')
  .with(null, () => 'it is null!')
  .with(undefined, () => 'it is undefined!')
  .exhaustive();

console.log(output);
// => 'it is either null or undefined!'

P.bigint wildcard

The P.bigint pattern will match any value of type bigint.

import { match, P } from 'ts-pattern';

const input = 20000000n;

const output = match<bigint | null>(input)
  .with(P.bigint, () => 'it is a bigint!')
  .otherwise(() => '?');

console.log(output);
// => 'it is a bigint!'

P.symbol wildcard

The P.symbol pattern will match any value of type symbol.

import { match, P } from 'ts-pattern';

const input = Symbol('some symbol');

const output = match<symbol | null>(input)
  .with(P.symbol, () => 'it is a symbol!')
  .otherwise(() => '?');

console.log(output);
// => 'it is a symbol!'

P.array patterns

To match on arrays of unknown size, you can use P.array(subpattern). It takes a sub-pattern, and will match if all elements in the input array match this sub-pattern.

import { match, P } from 'ts-pattern';

type Input = { title: string; content: string }[];

let input: Input = [
  { title: 'Hello world!', content: 'This is a very interesting content' },
  { title: 'Bonjour!', content: 'This is a very interesting content too' },
];

const output = match(input)
  .with(
    P.array({ title: P.string, content: P.string }),
    (posts) => 'a list of posts!'
  )
  .otherwise(() => 'something else');

console.log(output);
// => 'a list of posts!'

Matching variadic tuples with P.array

In TypeScript, Variadic Tuple Types are array types created with the ... spread operator, like [string, ...string[]], [number, ...boolean[], string] etc. You can match against variadic tuple types using array literals containing ...P.array(subpattern):

import { match, P } from 'ts-pattern';

type Input = (number | string)[];

declare const input: Input;

const output = match(input)
  // P.array's parameter is optional
  .with([P.string, ...P.array()], (input) => input) // input: [string, ...(number | string)[]]
  .with(['print', ...P.array(P.string)], (input) => input) // input: ['print', ...string[]]
  // you can put patterns on either side of `...P.array()`:
  .with([...P.array(P.string), 'end'], (input) => input) // input: [...string[], 'end']
  .with(['start', ...P.array(P.string), 'end'], (input) => input) // input: ['start', ...string[], 'end']
  .otherwise((input) => input);

P.set patterns

To match a Set, you can use P.set(subpattern). It takes a sub-pattern, and will match if all elements inside the set match this sub-pattern.

import { match, P } from 'ts-pattern';

type Input = Set<string | number>;

const input: Input = new Set([1, 2, 3]);

const output = match(input)
  .with(P.set(1), (set) => `Set contains only 1`)
  .with(P.set(P.string), (set) => `Set contains only strings`)
  .with(P.set(P.number), (set) => `Set contains only numbers`)
  .otherwise(() => '');

console.log(output);
// => "Set contains only numbers"

P.map patterns

To match a Map, you can use P.map(keyPattern, valuePattern). It takes a subpattern to match against the key, a subpattern to match agains the value, and will match if all elements inside this map match these two sub-patterns.

import { match, P } from 'ts-pattern';

type Input = Map<string, string | number>;

const input: Input = new Map([
  ['a', 1],
  ['b', 2],
  ['c', 3],
]);

const output = match(input)
  .with(P.map(P.string, P.number), (map) => `map's type is Map<string, number>`)
  .with(P.map(P.string, P.string), (map) => `map's type is Map<string, string>`)
  .with(
    P.map(P.union('a', 'c'), P.number),
    (map) => `map's type is Map<'a' | 'c', number>`
  )
  .otherwise(() => '');

console.log(output);
// => "map's type is Map<string, number>"

P.when patterns

P.when lets you define your own logic to check if the pattern should match or not. If the predicate function given to when returns a truthy value, then the pattern will match for this input.

Note that you can narrow down the type of your input by providing a Type Guard function to P.when.

import { match, P } from 'ts-pattern';

type Input = { score: number };

const output = match({ score: 10 })
  .with(
    {
      score: P.when((score): score is 5 => score === 5),
    },
    (input) => '😐' // input is inferred as { score: 5 }
  )
  .with({ score: P.when((score) => score < 5) }, () => '😞')
  .with({ score: P.when((score) => score > 5) }, () => '🙂')
  .run();

console.log(output);
// => '🙂'

P.not patterns

P.not lets you match on everything but a specific value. it's a function taking a pattern and returning the opposite pattern.

import { match, P } from 'ts-pattern';

type Input = boolean | number;

const toNumber = (input: Input) =>
  match(input)
    .with(P.not(P.boolean), (n) => n) // n: number
    .with(true, () => 1)
    .with(false, () => 0)
    .run();

console.log(toNumber(2));
// => 2
console.log(toNumber(true));
// => 1

P.select patterns

P.select lets you pick a piece of your input data-structure and injects it in your handler function.

It's especially useful when pattern matching on deep data structure to avoid the hassle of destructuring it in the handler function.

Selections can be either named (with P.select('someName')) or anonymous (with P.select()).

You can have only one anonymous selection by pattern, and the selected value will be directly inject in your handler as first argument:

import { match, P } from 'ts-pattern';

type Input =
  | { type: 'post'; user: { name: string } }
  | { ... };

const input: Input = { type: 'post', user: { name: 'Gabriel' } }

const output = match(input)
    .with(
      { type: 'post', user: { name: P.select() } },
      username => username // username: string
    )
    .otherwise(() => 'anonymous');

console.log(output);
// => 'Gabriel'

If you need to select several things inside your input data structure, you can name your selections by giving a string to P.select(<name>). Each selection will be passed as first argument to your handler in an object.

import { match, P } from 'ts-pattern';

type Input =
  | { type: 'post'; user: { name: string }, content: string }
  | { ... };

const input: Input = { type: 'post', user: { name: 'Gabriel' }, content: 'Hello!' }

const output = match(input)
    .with(
      { type: 'post', user: { name: P.select('name') }, content: P.select('body') },
      ({ name, body }) => `${name} wrote "${body}"`
    )
    .otherwise(() => '');

console.log(output);
// => 'Gabriel wrote "Hello!"'

You can also pass a sub-pattern to P.select if you want it to only select values which match this sub-pattern:

type User = { age: number; name: string };
type Post = { body: string };
type Input = { author: User; content: Post };

declare const input: Input;

const output = match(input)
  .with(
    {
      author: P.select({ age: P.number.gt(18) }),
    },
    (author) => author // author: User
  )
  .with(
    {
      author: P.select('author', { age: P.number.gt(18) }),
      content: P.select(),
    },
    ({ author, content }) => author // author: User, content: Post
  )
  .otherwise(() => 'anonymous');

P.optional patterns

P.optional(subpattern) lets you annotate a key in an object pattern as being optional, but if it is defined it should match a given sub-pattern.

import { match, P } from 'ts-pattern';

type Input = { key?: string | number };

const output = match(input)
  .with({ key: P.optional(P.string) }, (a) => {
    return a.key; // string | undefined
  })
  .with({ key: P.optional(P.number) }, (a) => {
    return a.key; // number | undefined
  })
  .exhaustive();

P.instanceOf patterns

The P.instanceOf function lets you build a pattern to check if a value is an instance of a class:

import { match, P } from 'ts-pattern';

class A {
  a = 'a';
}
class B {
  b = 'b';
}

type Input = { value: A | B };

const input: Input = { value: new A() };

const output = match(input)
  .with({ value: P.instanceOf(A) }, (a) => {
    return 'instance of A!';
  })
  .with({ value: P.instanceOf(B) }, (b) => {
    return 'instance of B!';
  })
  .exhaustive();

console.log(output);
// => 'instance of A!'

P.union patterns

P.union(...subpatterns) lets you test several patterns and will match if one of these patterns do. It's particularly handy when you want to handle some cases of a union type in the same code branch:

import { match, P } from 'ts-pattern';

type Input =
  | { type: 'user'; name: string }
  | { type: 'org'; name: string }
  | { type: 'text'; content: string }
  | { type: 'img'; src: string };

declare const input: Input;

const output = match(input)
  .with({ type: P.union('user', 'org') }, (userOrOrg) => {
    // userOrOrg: User | Org
    return userOrOrg.name;
  })
  .otherwise(() => '');

P.intersection patterns

P.intersection(...subpatterns) lets you ensure that the input matches all sub-patterns passed as parameters.

class A {
  constructor(public foo: 'bar' | 'baz') {}
}

class B {
  constructor(public str: string) {}
}

type Input = { prop: A | B };

declare const input: Input;

const output = match(input)
  .with(
    { prop: P.intersection(P.instanceOf(A), { foo: 'bar' }) },
    ({ prop }) => prop.foo // prop: A & { foo: 'bar' }
  )
  .with(
    { prop: P.intersection(P.instanceOf(A), { foo: 'baz' }) },
    ({ prop }) => prop.foo // prop: A & { foo: 'baz' }
  )
  .otherwise(() => '');

P.string predicates

P.string has a number of methods to help you match on specific strings.

P.string.startsWith

P.string.startsWith(str) matches strings that start with the provided string.

const fn = (input: string) =>
  match(input)
    .with(P.string.startsWith('TS'), () => '🎉')
    .otherwise(() => '❌');

console.log(fn('TS-Pattern')); // logs '🎉'

P.string.endsWith

P.string.endsWith(str) matches strings that end with the provided string.

const fn = (input: string) =>
  match(input)
    .with(P.string.endsWith('!'), () => '🎉')
    .otherwise(() => '❌');

console.log(fn('Hola!')); // logs '🎉'

P.string.minLength

P.string.minLength(min) matches strings with at least min characters.

const fn = (input: string) =>
  match(input)
    .with(P.string.minLength(2), () => '🎉')
    .otherwise(() => '❌');

console.log(fn('two')); // logs '🎉'

P.string.maxLength

P.string.maxLength(max) matches strings with at most max characters.

const fn = (input: string) =>
  match(input)
    .with(P.string.maxLength(5), () => '🎉')
    .otherwise(() => 'too long');

console.log(fn('is this too long?')); // logs 'too long'

P.string.includes

P.string.includes(str) matches strings that contain the provided substring.

const fn = (input: string) =>
  match(input)
    .with(P.string.includes('!'), () => '✅')
    .otherwise(() => '❌');

console.log(fn('Good job! 🎉')); // logs '✅'

P.string.regex

P.string.regex(RegExp) matches strings if they match the provided regular expression.

const fn = (input: string) =>
  match(input)
    .with(P.string.regex(/^[a-z]+$/), () => 'single word')
    .otherwise(() => 'other strings');

console.log(fn('gabriel')); // logs 'single word'

P.number and P.bigint predicates

P.number and P.bigint have several of methods to help you match on specific numbers and bigints.

P.number.between

P.number.between(min, max) matches numbers between min and max.

const fn = (input: number) =>
  match(input)
    .with(P.number.between(1, 5), () => '✅')
    .otherwise(() => '❌');

console.log(fn(3), fn(1), fn(5), fn(7)); // logs '✅ ✅ ✅ ❌'

P.number.lt

P.number.lt(max) matches numbers smaller than max.

const fn = (input: number) =>
  match(input)
    .with(P.number.lt(7), () => '✅')
    .otherwise(() => '❌');

console.log(fn(2), fn(7)); // logs '✅ ❌'

P.number.gt

P.number.gt(min) matches numbers greater than min.

const fn = (input: number) =>
  match(input)
    .with(P.number.gt(7), () => '✅')
    .otherwise(() => '❌');

console.log(fn(12), fn(7)); // logs '✅ ❌'

P.number.lte

P.number.lte(max) matches numbers smaller than or equal to max.

const fn = (input: number) =>
  match(input)
    .with(P.number.lte(7), () => '✅')
    .otherwise(() => '❌');

console.log(fn(7), fn(12)); // logs '✅ ❌'

P.number.gte

P.number.gte(min) matches numbers greater than or equal to min.

const fn = (input: number) =>
  match(input)
    .with(P.number.gte(7), () => '✅')
    .otherwise(() => '❌');

console.log(fn(7), fn(2)); // logs '✅ ❌'

P.number.int

P.number.int() matches integers.

const fn = (input: number) =>
  match(input)
    .with(P.number.int(), () => '✅')
    .otherwise(() => '❌');

console.log(fn(12), fn(-3.141592)); // logs '✅ ❌'

P.number.finite

P.number.finite() matches all numbers except Infinity and -Infinity.

const fn = (input: number) =>
  match(input)
    .with(P.number.finite(), () => '✅')
    .otherwise(() => '❌');

console.log(fn(-3.141592), fn(Infinity)); // logs '✅ ❌'

P.number.positive

P.number.positive() matches positive numbers.

const fn = (input: number) =>
  match(input)
    .with(P.number.positive(), () => '✅')
    .otherwise(() => '❌');

console.log(fn(7), fn(-3.141592)); // logs '✅ ❌'

P.number.negative

P.number.negative() matches negative numbers.

const fn = (input: number) =>
  match(input)
    .with(P.number.negative(), () => '✅')
    .otherwise(() => '❌');

console.log(fn(-3.141592), fn(7)); // logs '✅ ❌'

Types

P.infer

P.infer<typeof somePattern> lets you infer a type of value from a type of pattern.

It's particularly useful when validating an API response.

const postPattern = {
  title: P.string,
  content: P.string,
  stars: P.number.between(1, 5).optional(),
  author: {
    firstName: P.string,
    lastName: P.string.optional(),
    followerCount: P.number,
  },
} as const;

type Post = P.infer<typeof postPattern>;

// posts: Post[]
const posts = await fetch(someUrl)
  .then((res) => res.json())
  .then((res: unknown): Post[] =>
    isMatching({ data: P.array(postPattern) }, res) ? res.data : []
  );

Although not strictly necessary, using as const after the pattern definition ensures that TS-Pattern infers the most precise types possible.

P.narrow

P.narrow<Input, typeof Pattern> will narrow the input type to only keep the set of values that are compatible with the provided pattern type.

type Input = ['a' | 'b' | 'c', 'a' | 'b' | 'c'];
const Pattern = ['a', P.union('a', 'b')] as const;

type Narrowed = P.narrow<Input, typeof Pattern>;
//     ^? ['a', 'a' | 'b']

Note that most of the time, the match and isMatching functions perform narrowing for you, and you do not need to narrow types yourself.

P.Pattern

P.Pattern<T> is the type of all possible pattern for a generic type T.

type User = { name: string; age: number };

const userPattern: Pattern<User> = {
  name: 'Alice',
};

Type inference

TS-Pattern takes advantage the most advanced features of TypeScript to perform type narrowing and accurate exhaustive matching, even when matching on complex data-structures.

Here are some examples of TS-Pattern's type inference features.

Type narrowing

When pattern-matching on a input containing union types, TS-Pattern will infer the most precise type possible for the argument of your handler function using the pattern you provide.

type Text = { type: 'text'; data: string };
type Img = { type: 'img'; data: { src: string; alt: string } };
type Video = { type: 'video'; data: { src: string; format: 'mp4' | 'webm' } };
type Content = Text | Img | Video;

const formatContent = (content: Content): string =>
  match(content)
    .with({ type: 'text' }, (text /* : Text */) => '<p>...</p>')
    .with({ type: 'img' }, (img /* : Img */) => '<img ... />')
    .with({ type: 'video' }, (video /* : Video */) => '<video ... />')
    .with(
      { type: 'img' },
      { type: 'video' },
      (video /* : Img | Video */) => 'img or video'
    )
    .with(
      { type: P.union('img', 'video') },
      (video /* : Img | Video */) => 'img or video'
    )
    .exhaustive();

When using P.select in a pattern, TS-Pattern will find and inject the selected value in your handler. The type of your handler's argument is inferred accordingly.

const formatContent = (content: Content): string =>
  match(content)
    .with(
      { type: 'text', data: P.select() },
      (content /* : string */) => '<p>...</p>'
    )
    .with(
      { type: 'video', data: { format: P.select() } },
      (format /* : 'mp4' | 'webm' */) => '<video ... />'
    )
    .with(
      { type: P.union('img', 'video'), data: P.select() },
      (data /* : Img['data'] | Video['data'] */) => 'img or video'
    )
    .exhaustive();

Type guard function

If you pass a type guard function to P.when, TS-Pattern will use its return type to narrow the input.

const isString = (x: unknown): x is string => typeof x === 'string';

const isNumber = (x: unknown): x is number => typeof x === 'number';

const fn = (input: { id: number | string }) =>
  match(input)
    .with({ id: P.when(isString) }, (narrowed /* : { id: string } */) => 'yes')
    .with({ id: P.when(isNumber) }, (narrowed /* : { id: number } */) => 'yes')
    .exhaustive();

Exhaustiveness checking

TS-Pattern will keep track of handled and unhandled cases of your input type. Even when pattern-matching on several union types at once, you only need to call .exhaustive() to make sure that all possible cases are correctly handled.

type Permission = 'editor' | 'viewer';
type Plan = 'basic' | 'pro';

const fn = (org: Plan, user: Permission): string =>
  match([org, user])
    .with(['basic', 'viewer'], () => {})
    .with(['basic', 'editor'], () => {})
    .with(['pro', 'viewer'], () => {})
    // Fails with `NonExhaustiveError<['pro', 'editor']>`
    // because the `['pro', 'editor']` case isn't handled.
    .exhaustive();

Want to learn how TS-Pattern is built?

Check out 👉 Type-Level TypeScript, an online course to learn how to take full advantage of the most advanced features of TypeScript!

Inspirations

This library has been heavily inspired by this great article by Wim Jongeneel: Pattern Matching in TypeScript with Record and Wildcard Patterns. It made me realize pattern matching could be implemented in userland and we didn't have to wait for it to be added to the language itself. I'm really grateful for that 🙏