microsoft / microsoft/TypeScript

Suggesion: Allow recursive/self-references in `satisfies` constraint

Open
#54,367 6 comments 5 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Domain: Something Else Possible Improvement
Dominant language
Go
Stars
111k
Forks
14.4k
Avg merge
1d 19h
Merged PRs (30d)
117

Description

Suggestion

🔍 Search Terms

satisfies keyword, recursion, recursive, constraint, self referential

✅ Viability Checklist

My suggestion meets these guidelines:

  • This wouldn't be a breaking change in existing TypeScript/JavaScript code

    It would only make previously rejected type assertions possible as it only widens the set of valid type restrictions.

  • This wouldn't change the runtime behavior of existing JavaScript code

    satisfies only exists during type checking and does not in generated JavaScript

  • This could be implemented without emitting different JS based on the types of the expressions

    As above

  • This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, new syntax sugar for JS, etc.)

    As above

  • This feature would agree with the rest of TypeScript's Design Goals.

    The same is already possible using a constrained identity function

⭐ Suggestion

It is currently not possible to reference the type of a variable that the satisfies constraint is applied on. Enabling this functionality should not be hard, as the same behavior can be recreated using a constrained identity function, see below. This would enable more precise verification of structures by the type checker.

📃 Motivating Example

// I define a (very simplified) state machine by the following type:
type States<S extends string> = Record<S, S>;

// I can then define a map of machine names to their states records:
const myMachines1 = {
  alan: {
    a: "b",
    b: "c",
    c: "c",
  },
  turing: {
    x: "y",
    y: "x",
  },
} satisfies Record<string, States<string>>;

// Problem: This is possible, though should be invalid:
const myMachines2 = {
  foo: {
    a: "x",
  },
  bar: {
    x: "a",
  }
} satisfies Record<string, States<string>>;

// Solution 1: Identity function with self-referential generic:
function defineMachines<T extends {
  [name in keyof T]: States<keyof T[name] & string>;
}>(machines: T) {
  return machines;
}

// Valid:
const myMachines3 = defineMachines({
  alan: {
    a: "b",
    b: "c",
    c: "c",
  },
  turing: {
    x: "y",
    y: "x",
  },
});

// Errors:
const myMachines4 = defineMachines({
  foo: {
    a: "x",
  },
  bar: {
    x: "a",
  }
});

// Solution 2: The same behavior should be achievable using the new `satisfies` keyword:
const myMachines5 = {
  foo: {
    a: "x",
  },
  bar: {
    x: "a",
  }
} satisfies {
  [name in keyof typeof myMachines5]: States<keyof typeof myMachines5[name] & string>;
};
// But errors: Block-scoped variable 'myMachines5' used before its declaration.

💻 Use Cases

As seen in the example above, this allows for a more precise control over (nested) type restrictions. It is already possible to self-reference type definitions, so why shouldn't this also be possible for variable declarations? The satisfies keyword does not modify the type itself to my knowledge.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the motivating satisfies examples and compare the rejected self-reference with the constrained identity-function solution. Determine how recursive constraints should behave for variable declarations, then validate that valid state-machine maps are accepted and invalid cross-references are rejected without changing runtime output.

Written by the indexing model from the issue text.

Assessment

Tech stack
typescript
Domain
compilers
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.