Suggestion: throws clause and typed catch clause

[nitzantomer]

The typescript type system is helpful in most cases, but it can’t be utilized when handling exceptions.
For example:

function fn(num: number): void {
    if (num === 0) {
        throw "error: can't deal with 0";
    }
}

The problem here is two fold (without looking through the code):

1. When using this function there’s no way to know that it might throw an error
2. It’s not clear what the type(s) of the error is going to be

In many scenarios these aren't really a problem, but knowing whether a function/method might throw an exception can be very useful in different scenarios, especially when using different libraries.

By introducing (optional) checked exception the type system can be utilized for exception handling.
I know that checked exceptions isn't agreed upon (for example Anders Hejlsberg), but by making it optional (and maybe inferred? more later) then it just adds the opportunity to add more information about the code which can help developers, tools and documentation.
It will also allow a better usage of meaningful custom errors for large big projects.

As all javascript runtime errors are of type Error (or extending types such as TypeError) the actual type for a function will always be type | Error.

The grammar is straightforward, a function definition can end with a throws clause followed by a type:

function fn() throws string { ... }
function fn(...) throws string | number { ... }

class MyError extends Error { ... }
function fn(...): Promise<string> throws MyError { ... }

When catching the exceptions the syntax is the same with the ability to declare the type(s) of the error:
catch(e: string | Error) { ... }

Examples:

function fn(num: number): void throws string {
    if (num === 0) {
        throw "error: can't deal with 0";
    }
}

Here it’s clear that the function can throw an error and that the error will be a string, and so when calling this method the developer (and the compiler/IDE) is aware of it and can handle it better.
So:

fn(0);

// or
try {
    fn(0); 
} catch (e: string) { ... }

Compiles with no errors, but:

try {
    fn(0); 
} catch (e: number) { ... }

Fails to compile because number isn't string.

Control flow and error type inference

try {
    fn(0);
} catch(e) {
    if (typeof e === "string") {
        console.log(e.length);
    } else if (e instanceof Error) {
        console.log(e.message);
    } else if (typeof e === "string") {
        console.log(e * 3); // error: Unreachable code detected
    }

    console.log(e * 3); // error: The left-hand side of an arithmetic operation must be of type 'any', 'number' or an enum type
}

function fn(num: number): void {
    if (num === 0) {
        throw "error: can't deal with 0";
    }
}

Throws string.

function fn2(num: number) {
    if (num < 0) {
        throw new MyError("can only deal with positives");
    }

    fn(num);
}

Throws MyError | string.
However:

function fn2(num: number) {
    if (num < 0) {
        throw new MyError("can only deal with positives");
    }

    try {
        fn(num);
    } catch(e) {
        if (typeof e === "string") {
           throw new MyError(e);
       } 
    }
}

Throws only MyError.

[DanielRosenwasser]

Just to clarify - one the ideas here is not to force users to catch the exception, but rather, to better infer the type of a catch clause variable?

[nitzantomer]

@DanielRosenwasser
Yes, users won't be forced to catch exceptions, so this is fine with the compiler (at runtime the error is thrown of course):

function fn() {
    throw "error";
}

fn();

// and
try {
    fn();
} finally {
    // do something here
}

But it will give developers a way to express which exceptions can be thrown (would be awesome to have that when using other libraries .d.ts files) and then have the compiler type guard the exception types inside the catch clause.

[zpdDG4gta8XKpMCd]

how is a checked throw different from Tried<Result, Error>?

type Tried<Result, Error> = Success<Result> | Failure<Error>;
interface Success<Result> { kind: 'result', result: Result } 
interface Failure<Error> { kind: 'failure', error: Error }
function isSuccess(tried: Tried<Result, Error>): tried is Success<Result> {
   return tried.kind === 'result';
}
function mightFail(): Tried<number, string> {
}
const tried = mightFail();
if (isSuccess(tried)) {
    console.log(tried.success);
}  else {
    console.error(tried.error);
}

instead of

try {
    const result: Result = mightFail();
    console.log(success);
} catch (error: Error) {
    console.error(error);
}

[nitzantomer]

@Aleksey-Bykov

You're suggesting not to use throw at all in my code and instead wrap the results (in functions that might error).
This approach has a few drawbacks:

• This wrapping creates more code
• It requires that all chain of invoked function return this wrapped value (or error) or alternatively the function that gets Tried<> can not choose to ignore the error.
• It is not a standard, 3rd party libraries and the native js throw errors

Adding throws will enable developers who choose to to handle errors from their code, 3rd libraries and native js.
As the suggestion also requests for error inferring, all generated definition files can include the throws clause.
It will be very convenient to know what errors a function might throw straight from the definition file instead of the current state where you need to go to the docs, for example to know which error JSON.parse might throw I need to go to the MDN page and read that:

Throws a SyntaxError exception if the string to parse is not valid JSON

And this is the good case when the error is documented.

[zpdDG4gta8XKpMCd]

And this is the good case when the error is documented.

is there a reliable way in javascript to tell apart SyntaxError from Error?

• yes, it's more code, but since a bad situation is represented in an object, it can be passed around to be processed, discarded, stored or transformed into a valid result just like any other value

• you can ignore tried by returning tried too, tried can be viewed as a monad, look for monadic computations

  function mightFail(): Tried<number, string> {
  }
  function mightFailToo(): Tried<number, string> {
       const tried = mightFail();
       if (isSuccess(tried))  { 
            return successFrom(tried.result * 2);
       } else {
            return tried;
       }
  }
  

• it's standard enough for your code, when it comes to 3rd party libs throwing an exception it generally means a gameover for you, because it is close to impossible to reliably recover from an exception, reason is that it can be thrown from anywhere inside the code terminating it at an arbitrary position and leaving its internal state incomplete or corrupt

• there is no support for checked exceptions from JavaScript runtime, and i am afraid it cannot be implemented in typescript alone

other than that encoding an exception as a special result case is a very common practice in FP world

whereas splitting a possible outcome into 2 parts:

• one delivered by the return statement and
• another delivered by throw

looks a made up difficulty

in my opinion, throw is good for failing fast and loud when nothing you can do about it, explicitly coded results are good for anything that implies a bad yet expected situation which you can recover from

[zpdDG4gta8XKpMCd]

consider:

// throw/catch
declare function doThis(): number throws string;
declare function doThat(): number throws string;
function doSomething(): number throws string {
    let oneResult: number | undefined = undefined;
    try {
        oneResult = doThis();
    } catch (e) {
        throw e;
    }

    let anotherResult: number | undefined = undefined;
    try {
        anotherResult = doThat();
    } catch (e) {
        throw e;
    }
    return oneResult + anotherResult;
}

// explicit results
declare function doThis(): Tried<number, string>;
declare function doThat(): Tried<number, string>;
function withBothTried<T, E, R>(one: Tried<T, E>, another: Tried<T, E>, haveBoth: (one: T, another: T) => R): Tried<T, R> {
    return isSuccess(one)
        ? isSuccess(another)
            ? successFrom(haveBoth(one.result, another.result))
            : another
        : one;
}
function add(one: number, another: number) { return one + another; }
function doSomething(): Tried<number, string> {
    return withBothTried(
        doThis(),
        doThat(),
        add
    );
}

[nitzantomer]

@Aleksey-Bykov

My point with JSON.parse might throwing SyntaxError is that I need to look the function up in the docs just to know that it might throw, and it would be easier to see that in the .d.ts.
And yes, you can know that it's SyntaxError with using instanceof.

You can represent the same bad situation with throwing an error.
You can create your own error class which extends Error and put all of the relevant data that you need in it.
You're getting the same with less code.

Sometimes you have a long chain of function invocations and you might want to deal with some of the errors in different levels of the chain.
It will be pretty annoying to always use wrapped results (monads).
Not to mention that again, other libraries and native errors might be thrown anyway, so you might end up using both monads and try/catch.

I disagree with you, in a lot of cases you can recover from thrown errors, and if the language lets you express it better than it will be easier to do so.

Like with a lot of things in typescript, the lack of support of the feature in javascript isn't an issue.
This:

try {
	mightFail();
} catch (e: MyError | string) {
	if (e instanceof MyError) { ... }
	else if (typeof e === "string") { ... }
	else {}
}

Will work as expected in javascript, just without the type annotation.

Using throw is enough to express what you're saying: if the operation succeeded return the value, otherwise throw an error.
The user of this function will then decide if he wants to deal with the possible errors or ignore them.
You can deal with only errors you thrown yourself and ignore the ones which are 3rd party for example.

[zpdDG4gta8XKpMCd]

if we talking about browsers instanceof is only good for stuff that originates from the same window/document, try it:

var child = window.open('about:blank');
console.log(child.Error === window.Error);

so when you do:

try { child.doSomething(); } catch (e) { if (e instanceof SyntaxError) { } }

you won't catch it

another problem with exceptions that they might slip into your code from far beyond of where you expect them to happen

try {
   doSomething(); // <-- uses 3rd party library that by coincidence throws SyntaxError too, but you don' t know it 
} catch (e) {}

[zpdDG4gta8XKpMCd]

besides instanceof is vulnerable to prototype inheritance, so you need to be extra cautions to always check against the final ancestor

class StandardError {}
class CustomError extends StandardError {
}
function doSomething() { throw new CustomError(); }
function oldCode() {
   try {
      doSomething();
   } catch (e) {
      if (e instanceof StandardError) {
          // problem
      }
   }
}

[gcnew]

@Aleksey-Bykov Explicitly threading errors as you suggest in monadic structures is quite hard and daunting task. It takes a lot of effort, makes the code hard to understand and requires language support / type-driven emit to be on the edge of being bearable. This is a comment comming from somebody who puts a lot of effort into popularising Haskell and FP as a whole.

It is a working alternative, especially for enthusiasts (myself included), however I don't think it's a viable option for the larger audience.

[aluanhaddad]

Actually, my main concern here is that people will start subclassing Error. I think this is a terrible pattern. More generally, anything that promotes the use of the instanceof operator is just going to create additional confusion around classes.

[zpdDG4gta8XKpMCd]

This is a comment comming from somebody who puts a lot of effort into popularising Haskell and FP as a whole.

i really think this should be pushed harder to the audience, not until it's digested and asked for more can we have better FP support in the language

and it's not as daunting as you think, provided all combinators are written already, just use them to build a data flow, like we do in our project, but i agree that TS could have supported it better: #2319

[gcnew]

Monad transformers are a real PITA. You need lifting, hoisting and selective running fairly often. The end result is hardly comprehendible code and much higher than needed barrier of entry. All the combinators and lifting functions (which provide the obligatory boxing/unboxing) are just noise distracting you from the problem at hand. I do believe that being explicit about state, effects, etc is a good thing, but I don't think we have found a convenient wrapping / abstraction yet. Until we find it, supporting traditional programming patterns seems like the way to go without stopping to experiment and explore in the mean time.

PS: I think we need more than custom operators. Higher Kinded Types and some sort of type classes are also essential for a practical monadic library. Among them I'd rate HKT first and type classes a close second. With all that said, I believe TypeScript is not the language for practicing such concepts. Toying around - yes, but its philosophy and roots are fundamentally distant for a proper seamless integration.

[gcnew]

Back to the OP question - instanceof is a dangerous operator to use. However explicit exceptions are not limited to Error. You can throw your own ADTs or custom POJO errors as well. The proposed feature can be quite useful and, of course, can also be misused pretty hard. In any case it makes functions more transparent which is undoubtedly a good thing. As a whole I'm 50/50 on it :)

[nitzantomer]

@Aleksey-Bykov

Developers should be aware of the different js issues you described, after all adding throws to typescript doesn't introduce anything new to js, it only gives typescript as a language the ability to express an existing js behavior.

The fact that 3rd party libraries ca throw errors is exactly my point.
If their definition files were to include that then I will have a way to know it.

@aluanhaddad
Why is it a terrible pattern to extend Error?

@gcnew
As for instanceof, that was just an example, I can always throw regular objects which have different types and then use type guards to differentiate between them.
It will be up to the developer to decide what type of errors he wishes to throw, and it probably is the case already, but currently there's no way to express that, which is what this suggestion wants to solve.

[gcnew]

@nitzantomer Subclassing native classes (Error, Array, RegExp, etc) was not supported in older ECMAScript versions (prior to ES6). The down level emit for these classes gives unexpected results (best effort is made but this is as far as one can go) and is the reason for numerous issues logged on daily basis. As a rule of thumb - don't subclass natives unless you are targeting recent ECMAScript versions and really know what you are doing.

[nitzantomer]

@gcnew
Oh, I'm well aware of that as I spent more than a few hours trying to figure out what went wrong.
But with the ability to do so now there shouldn't be a reason not to (when targeting es6).

In anycase this suggestion doesn't assume that the user is subclassing the Error class, it was just an example.

[gcnew]

@nitzantomer I'm not arguing that the suggestion is limited to Error. I just explained why it's a bad pattern to subclass it. In my post I actually defended the stance that custom objects or discriminated unions may be used as well.

instanceof is dangerous and considered an anti-pattern even if you take out the specificities of JavaScript - e.g. Beware of instanceof operator. The reason is that the compiler cannot protect you against bugs introduced by new subclasses. Logic using instanceof is fragile and does not follow the open/closed principle, as it expects only a handful of options. Even if a wildcard case is added, new derivates are still likely to cause errors as they may break assumptions made at the time of writing.

For the cases where you want to distinguish among known alternatives TypeScript has Tagged Unions (also called discriminated unions or algebraic data types). The compiler makes sure that all cases are handled which gives you nice guarantees. The downside is that if you want to add a new entry to the type, you'll have to go through all the code discriminating on it and handle the newly added option. The upside is that such code would have most-likely been broken, but would have failed at runtime.

[gcnew]

I just gave this proposal a second thought and became against it. The reason is that if throws declarations were present on signatures but were not enforced, they can already be handled by documentation comments. In the case of being enforced, I share the sentiment that they'd become irritating and swallowed fast as JavaScript lacks Java's mechanism for typed catch clauses. Using exceptions (especially as control flow) has never been an established practice as well. All of this leads me to the understanding that checked exceptions bring too little, while better and presently more common ways to represent failure are available (e.g. union return).

[nitzantomer]

@gcnew
This is how it's done in C#, the problem is that docs aren't as standard in typescript.
I do not remember coming across a definition file which is well documented. The different lib.d.ts files do contain comments, but those do not contain thrown errors (with one exception: lib.es6.d.ts has one throws in Date[Symbol.toPrimitive](hint: string)).

Also, this suggestion takes error inferring into account, something that won't happen if errors are coming from documentation comments. With inferred checked exceptions the developer won't even need to specify the throws clause, the compiler will infer it automatically and will use it for compilation and will add it to the resulting definition file.

I agree that enforcing error handling isn't a good thing, but having this feature will just add more information which can be then used by those who wish to.
The problem with:

... there are better and presently more common ways to represent failure

Is that there's no standard way of doing it.
You might use union return, @Aleksey-Bykov will use Tried<>, and a developer of another 3rd party library will do something completely different.
Throwing errors is a standard across languages (js, java, c#...) and as it's part of the system and not a workaround, it should (in my opinion) have better handling in typescript, and a proof of that is the number of issues I've seen here over time which ask for type annotation in the catch clause.

[HolgerJeromin]

I would love to have information in the tooltip in VS if a function (or called function) can throw. For *.d.ts files we probably need a fake parameter like this since TS2.0.

[nitzantomer]

@HolgerJeromin
Why would it be needed?

[zpdDG4gta8XKpMCd]

here is a simple question, what signature should be inferred for dontCare in the code below?

function mightThrow(): void throws string {
   if (Math.random() > 0.5) {
       throw 'hey!';
   }
}

function dontCare() {
   return mightThrow();
}

according to what you said in your proposal it should be

function dontCare(): void throws string {

i say it should be a type error since a checked exception wasn't properly handled

function dontCare() { // <-- Checked exception wasn't handled.
         ^^^^^^^^^^

why is that?

because otherwise there is a very good chance of getting the state of the immediate caller corrupt:

class MyClass {
    private values: number[] = [];

    keepAllValues(values: number[]) {
       for (let index = 0; index < values.length; index ++) {
            this.values.push(values[index]); 
            mightThrow();
       }
    }
}

if you let an exception to slip through you can not infer it as checked, because the behavior contract of keepAllValues would be violated this way (not all values were kept despite the original intent)

the only safe way to is catch them immediately and rethrow them explicitly

    keepAllValues(values: number[]) {
           for (let index = 0; index < values.length; index ++) {
                this.values.push(values[index]); 
                try {
                    mightThrow();
                } catch (e) {
                    // the state of MyClass is going to be corrupt anyway
                    // but unlike the other example this is a deliberate choice
                    throw e;
                }
           }
    }

otherwise despite the callers know what can be trown you can't give them guarantees that it's safe to proceed using code that just threw

so there is no such thing as automatic checked exception contract propagation

and correct me if i am wrong, this is exactly what Java does, which you mentioned as an example earlier

[nitzantomer]

@Aleksey-Bykov
This:

function mightThrow(): void {
   if (Math.random() > 0.5) {
       throw 'hey!';
   }
}

function dontCare() {
   return mightThrow();
}

Means that both mightThrow and dontCare are inferred to throws string, however:

function dontCare() {
    try {
        return mightThrow();
    } catch (e: string) {
        // do something
    }
}

Won't have a throw clause because the error was handled.
This:

function mightThrow(): void throws string | MyErrorType { ... }

function dontCare() {
    try {
        return mightThrow();
    } catch (e: string | MyErrorType) {
        if (typeof e === "string") {
            // do something
        } else { throw e }
    }
}

Will have throws MyErrorType.

As for your keepAllValues example, I'm not sure what you mean, in your example:

class MyClass {
    private values: number[] = [];

    keepAllValues(values: number[]) {
       for (let index = 0; index < values.length; index ++) {
            this.values.push(values[index]); 
            mightThrow();
       }
    }
}

MyClass.keepAllValues will be inferred as throws string because mightThrow might throw a string and that error was not handled.

[kaleidawave]

I have a few insights so thought I would write some here.

The first is what benefits having a type defined for the err variable in a catch would bring:

• Being able to understand when statements can throw. Given getConfig() it would be nice to get feedback in the editor whether it throws a ParseError | InvalidConfig or a NetworkError which would help with what handling should do. Currently figuring that is left to reading source and reading documentation, which is what existing typing support aims to reduce.
• type safety within the source, this is a good goal but given all the other loopholes TS has this doesn't complete or finish the puzzle.

Inference of thrown values

The big problem as noted is that this requires adding / changing pretty much all function type descriptors. Ones that aren't changed would have to assume some value.

IMO to do this properly and today TS needs to able to recognise thrown types from functions. To do this TS would need to add an 'events system'. When synthesizing a function, it would need to record all throw events. Handing throw statements is trivial, the hard part is function calls (and as mentioned that getters) as they can throw from there invocation. TS would need to record values of parameters etc, it all gets quite complex. The hard parts that code from throw in an inferred events system.

• Proxy, objects after Object.defineProperty, etc
• Generators
• Calling any-ish types
• What internal functions (standard library, node, browser) throw
• What internal functions access and call. e.g. customElements.define("...", class A {}, { get extends() { throw "Extends read!" } })
• non synchronous calls, e.g. the function argument to queueMicrotask

This would be a big change to TS's direction, moving away from type annotations as the source of truth instead to code as truth...

---

Another part that I haven't seen mentioned, is not just registering throw, but also errors emitted from the JS engine itself (aka TypeError and ReferenceError). Undefined variables, getting properties on a null are cases. Currently they are a hard type error. Could that be relaxed, rather than being a error at the call site it would be an error that it is uncaught by a catch? e.g. could you read .a on { a: string } | { b: number }.

---

The checked exceptions is also interesting task. For some contexts like the top level, uncaught errors would be useful to be reported. Functions on the other hand there are uses cases for errors to bubble up. For the callback for setTimeout it would be good to make sure it doesn't throw / is handled, maybe it could take a PureFunction instead?

---

Finally, while I do like Rust's Result structure, I think JS is missing a lot of the infrastructure for the using { ok: T } | { err: E } pattern (notably match and ?). I can't see a increase in the result pattern coming soon as would require changing browser APIs and moving away from a huge part of the specification...

[zefir-git]

@kaleidawave in my view a function's throws type could be implemented the same way as its regular returns type. I do not see why events system would be needed. Just like a return statement, an uncaught throw statement terminates a function. TypeScript is already able to detect return types when calling methods (and getters too).

Example:

class A {
    public foo?: string;
    public get bar() {
        return this.foo ?? "was undefined";
    }
    public baz() {
        return this.bar;
    }
}

const a = new A().baz(); // type detected as `string`

https://tsplay.dev/weaLKw

Therefore, I do not see how this would be much different:

class A {
    public foo?: Error;
    public get bar() {
        throw this.foo ?? new TypeError("was undefined");
    }
    public baz() {
        return this.bar;
    }
}
try {
    new A().baz();
}
catch (e) {
    // e: Error | TypeError
}

Could be further simplified by adding optional throws to functions (just like specifying the return type can be optional).

[titouandk]

This is a huge problem in JS/TS.

Some libraries throws exceptions as a normal part of the flow – as a sentinel value if you wish.
Developers of those libs expects you to catch those informative exceptions.

The problem? You are not even informed of their existence.

A concrete example:

You would expect the following code (copying source into destination) to fail silently if the file already exists in destination – as the linux command cp -n source destination does:

await vscode.workspace.fs.copy(source, destination, { overwrite: false });

But instead, in this case, this function will throw an exception to inform you that the file already exists at destination!

Yes. Indeed. It will throw an exception to let you know that everything is fine! And you are not even warned of this possible behavior by the type system!

This is hell 🤪

JS devs do not deserve to get PTSD-like symptoms each time they call a function – worrying that it may explode to their face at any time, without warning.

[thelinuxlich]

throws should work solely for functions that were typed to use it, none more, none less.

[galaxynova1999]

why is this issue closed? Is there any definitive conclusion?

[ericbf]

Just scroll up. #13219 (comment)

[KashubaK]

I love the idea of checked exceptions. Personally I see value in this in regards to our team projects. If I implement function X that throws errors, and a junior developer wants to use it, they should handle the required errors. If TS caught them, I wouldn't have to rely on having to flag it down in code review (or if someone else was conducting the review, hoping the reviewer knew about the errors!)

As noted, annotating function signatures sounds like it could be pretty nutty. I wonder if this information could be placed alongside throw statements with new safe/unsafe keywords, for example:

async function doSomething(arg1: string, arg2: number) {
  if (arg1.length > 10) {
    throw safe new StringTooLongError();
  }
  
  if (arg2 < 0) {
    throw safe new NumberLessThanZeroError();
  }
  
  // `safe` meaning "You did it wrong" error, don't enforce try/catch
 
  // However if you think the developer absolutely should be required to try/catch...
  
  try {
    return await fetch('http://some-api.com/');
  } catch (err) {
    // fetch can throw "TypeError" (i.e. 'Failed to fetch')
    // If the fetch types were updated, it could even force the developer to handle it.
    // Regardless, if the developer knows of an undocumented, unsafe error, they could make it unsafe by proxy:
    throw unsafe new RequestFailedError();
  }
  
  // Non-annotated errors behave as-is:
  throw new Error('Whoops!')
}

In this case, the resulting type signature would be something like:

function doSomething(arg1: string, arg2: number): Promise<Response> throws unsafe RequestFailedError | safe StringTooLongError | safe NumberLessThanZeroError;

Example usage:

await doSomething("hello", 5);

// TS error: unsafe RequestFailedError exception must be caught

try {
  await doSomething("hello", 5);
} catch (err) { // TS error: unsafe RequestFailedError exception not narrowed in catch block
  // err: RequestFailedError | StringTooLongError | NumberLessThanZeroError
  // Problem: after handling all of the above, does err become never, or unknown?
  // Ideally it would become unknown, but current union behavior would reflect never.
}

In the case of an un-documented error, to avoid unintentional inference, they should be ignored completely. throw new UnknownError() should not be included in throws annotations. This ensures that if a developer upgrades their TS package, they don't get hit with 1,000+ type errors requiring a bunch of copy/paste try {} catch {}. Though perhaps a strictErrors option could be opted-out-of in tsconfig.json.

I think this results in a much clearer experience. Developers opt into annotation, explicitly choosing which exceptions need to be handled and which ones don't. Or avoid it entirely if undesirable, however I imagine library consumers would argue otherwise. A counter argument to this may be "If we don't enforce annotation, we're stuck with the same problem that exists now with the lack of documentation." My response to this is, by providing rails (i.e. the "right way") then maintainers have to think less. I'm more likely to add one-word "unsafe" to my code, as opposed to maintaining the alternative in some sort of API documentation (that may/may not exist in the first place.)

This leads into a point regarding one of the reconsideration points:

Widespread adoption and documentation of strong exception hierarchies in JS libraries in the wild

In my opinion, TypeScript is best positioned to champion this effort with this feature. I don't believe, without some sort of innovation, that this could be achieved.

The idea here is incremental adoption, to encourage developers (and library maintainers) to handle errors more elegantly. Just because the culture isn't prevalent doesn't mean it can't be developed over time. One of the biggest complaints about JS is it's the "wild-west" where anything could go wrong, and this would be a good step in the right direction.

To address the point of:

Any feature here implies a huge amount of new information in .d.ts files that isn't documented in the first place

I think that's one of the main reasons developers want this feature. If this were to be implemented, I'm sure we'd find many consumers asking maintainers to "annotate exceptions in X function". This working out-of-box for all existing TS projects is not feasible, however given time for adoption, the ecosystem could vastly improve.

Constructing a RegExp from non-hardcoded input is so vanishingly rare that it seems obnoxious to force a try/catch around new RegExp("[A-Z]") since that technically "can" throw an unavoidable exception, even though by inspection it clearly doesn't.

I wonder if there could be some sort of way to bypass this.

// RegExp constructor is unsafe by default
const regex =  new RegExp("[A-Z]") as safe;

This sounds scary, but various linter rules could be implemented to push developers away from using as safe if desired.

Alternatively, a function could work around this if implementing as safe in TS is not feasible:

function safe<T>(fn: () => T throws unsafe unknown): T {
  try {
    return fn()
  } catch (error) {
    throw safe error;
  }
}

const regex = safe(() => new RegExp("[a-z]"));

Which less ideal, but more readable than having to do try/catch manually everytime.

---

Forgive me if any of these suggestions are naive or lack understanding! This feature seems very ideal to me and I'd rather the conversation continue than otherwise. In closing, my two-cents would be:

• Leave undocumented errors as-is
• Encourage new code to document errors by developer demand
• Allow developers to opt-out of checked exceptions via a strict option
• Make it easy to document errors to mitigate required effort
• Incremental adoption is better than none

[gastonmorixe]

+1 I think consensus seems to be we are not requesting a perfect solution, a good one would be a great first step.

This is a key place TS can help to reduce bugs and improve developers' efficiency to al least warn them against some known possible errors. It shouldn't be that hard.

I liked @KashubaK safe and unsafe idea as well.

Here's the swift official doc about throwing functions
https://docs.swift.org/swift-book/documentation/the-swift-programming-language/errorhandling/

and a good example

class VendingMachine {
    var inventory = [
        "Candy Bar": Item(price: 12, count: 7),
        "Chips": Item(price: 10, count: 4),
        "Pretzels": Item(price: 7, count: 11)
    ]
    var coinsDeposited = 0


    func vend(itemNamed name: String) throws {
        guard let item = inventory[name] else {
            throw VendingMachineError.invalidSelection
        }


        guard item.count > 0 else {
            throw VendingMachineError.outOfStock
        }


        guard item.price <= coinsDeposited else {
            throw VendingMachineError.insufficientFunds(coinsNeeded: item.price - coinsDeposited)
        }


        coinsDeposited -= item.price


        var newItem = item
        newItem.count -= 1
        inventory[name] = newItem


        print("Dispensing \(name)")
    }
}

[KashubaK]

On the point of the TS team's desire for an existing adoption of some sort of exception handling standard, I wonder what that could look like. Perhaps some other CLI tool that looks at @throws JSDoc tags? There was some mention of this in #31329, however I do agree that it's out of scope for TS to type check based off JSDoc comments.

If JSDoc and new TS syntax is out of the question for now, then the alternative is some sort of descriptor file that documents a file's exception handling.

(To be clear, I'm not suggesting TS support something like this. I'm spitballing an idea for a different tool.)

For example:

src/VendingMachine.ts

export class VendingMachine {
  inventory = {
    "Candy Bar": new Item({ price: 12, count: 7 }),
    "Chips": new Item({ price: 10, count: 4 }),
    "Pretzels": new Item({ price: 7, count: 11 })
  }
  
  coinsDeposited = 0;

  vend(name: string) {
    const item = inventory[name];

    if (!item) {
      throw new InvalidSelectionError();
    }

    if (item.count === 0) {
      throw new OutOfStockError();
    }

    if (item.price > this.coinsDeposited) {
      throw new InsufficientFundsError({ coinsNeeded: item.price - this.coinsDeposited })
    }

    this.coinsDeposited -= item.price;
    item.count -= 1;

    console.log(`Dispensing ${name}`)
  }
}

Then an adjacent file that documents this:

src/VendingMachine.errors.ts

export default {
  VendingMachine: {
    vend: {
      safe: { // Could accept an object containing more documentation
        error: InvalidSelectionError,
        description: "Lorem ipsum sit dolor amet",
      },
      unsafe: [ // If there are multiple errors, use an array
        OutOfStockError, // For ease of use, accept the error class without extra documentation
        { error: InsufficientFundsError, description: "coins deposited is less than the item price" }
      ]
    }
  }
}

Then have a CLI tool that introspects this information and inlines errors/warnings where applicable. I don't know how feasible this would be, or if there are improvements that could be made.

However I think if we're passionate about this feature, we should try to address the TS team's concerns in one way or another. If some sort of solution could be built and there is sufficient developer demand for adoption then a TS implementation could be warranted in the future.

[kvenn]

@KashubaK, riffing off that same idea, I thought about creating an ESLint rule since that's already a widely adopted standard. It has some serious limitations, but maybe ones that someone who has built a rule before could weigh in on.

The benefits would be that, for anyone using ESLint, it would be just one line they'd need to add to their config (or two, if we made multiple rules).

I took a pass at a convo with ChatGPT which I'll share here in case it has some merits: https://chat.openai.com/share/8354ae88-a968-41c8-ab9a-eec8a23fa5ef

The requirement I gave it:

/**
 * @throws {@link TypeError} if the notification is not found
 */
function calleeFunction1() {
	
}

/**
 * This doc has text on top
 * @throws {@link TypeError} if the notification is not found
 */
function calleeFunction2() {

}

/**
 * @throws {@link TypeError} if the notification is not found
 * This doc has text on bottom
 */
function calleeFunction3() {

}

function callerFunction() {
	// This line should have a lint warning because it isn't wrapped in try / catch
	calleeFunction1()
	
	// So should this one
	calleeFunction2()
	
	// So should this one
	calleeFunction3()
}

The relevant es-lint rule it created

module.exports = {
  rules: {
    "enforce-exception-handling": {
      create: function(context) {
        const sourceCode = context.getSourceCode();
        let functionsThatThrow = new Set();
        let scopeChain = [];

        function enterScope(node) {
          scopeChain.push(node);
        }

        function exitScope() {
          scopeChain.pop();
        }

        function currentScope() {
          return scopeChain.map(node => sourceCode.getText(node)).join(' ');
        }

        return {
          Program: enterScope,
          FunctionDeclaration: enterScope,
          FunctionExpression: enterScope,
          ArrowFunctionExpression: enterScope,
          "Program:exit": exitScope,
          "FunctionDeclaration:exit": exitScope,
          "FunctionExpression:exit": exitScope,
          "ArrowFunctionExpression:exit": exitScope,

          FunctionDeclaration(node) {
            const throwsComment = node.leadingComments &&
              node.leadingComments.find(comment => comment.value.includes("@throws"));

            if (throwsComment) {
              functionsThatThrow.add(currentScope() + ' ' + node.id.name);
            }
          },

          CallExpression(node) {
            const functionId = currentScope() + ' ' + node.callee.name;

            if (functionsThatThrow.has(functionId)) {
              let parent = node;

              while (parent) {
                if (parent.type === "TryStatement") return;
                parent = parent.parent;
              }

              context.report({
                node,
                message: "The function " + node.callee.name + " can throw an exception and needs to be called inside a try-catch block."
              });
            }
          }
        };
      }
    }
  }
};

As a side note, I agree with everything you've said. This would be a MASSIVE level-up to TypeScript and I'm sure we could find a way to make this feature opt-in for those who value it. And I bet we'd all be surprised how quickly it's adopted (by large library authors). Function calls into libraries that throw undocumented errors (even though they explicitly call throw) has been a real struggle with choosing node as my server solution.

[KashubaK]

@kvenn I wonder if that could also be implemented in a Language Service Plugin?

As for actually emitting errors, we could adopt a strategy akin to how tsc-strict has a CLI that will actually throw them (for CI or other purposes)

One reason I'm leaning away from relying on @throws is the lack of segregation between safe and unsafe errors

[KashubaK]

I'm kind of liking the idea of a JSDoc linter rule solution, actually. Ideally it could:

• Warn undocumented throw statements
• Error on uncaught "unsafe" exceptions
• Warn on uncaught "safe" exceptions

In order to make this impactful a significant effort will be required.

1. Implement a JSDoc based linter rule (arguably the easiest step in this process.)
2. Advocate for popular libraries to document @throws in their code
   • This might look like "require future PRs to add @throws in new code" to start
3. Listen for push-back and address concerns
4. Submit PRs that adds this documentation to help get maintainers on board
5. Advocate for boilerplate generators (yarn create vite, create-react-app, etc.) to include our linter rule

Perhaps even go over to DefinitelyTyped and elsewhere to add documentation for libraries that we personally use.

I also wonder how the TS team would respond to incrementally updating the lib types to include @throws: https://github.com/microsoft/TypeScript/tree/main/src/lib

Considering the 1,400+ upvotes on this issue, if we had a solid plan we could make this work!

[gastonmorixe]

@RyanCavanaugh would it be possible to reopen this please? Thank you

[KashubaK]

I don't think there's any actionable items in regards to this specific issue at this time. I think the next step should revolve around implementing a solution that encourages error checking in hopes to create widespread adoption of error documentation. After that point I imagine this issue could be brought back up, though I am definitely curious about their thoughts on our recent comments.

I'm going to try my hand at a linter rule over the weekend. If it goes well, I'll link the repo back here and we can move that conversation there.

[gastonmorixe]

I think it’s important enough to not be closed and encourage more discussion

[kvenn]

Agreed @KashubaK, I think the TypeScript team has made their opinion known. While I personally don't agree with the decision to deprioritize exception handling, I understand it. They don't think people will adopt it, and so it's on the community to prove otherwise. (That being said, I'd love if they reconsidered :))

Their reconsideration points made that clear:

Reconsideration Points
What would change our evaluation here? Two primary things come to mind:

1. Widespread adoption and documentation of strong exception hierarchies in JS libraries in the wild
2. TC39 proposals to implement some sort of pattern-matching criteria to catch exceptions (arguably this would just work "by itself" the same way instanceof works today)

@KashubaK, your steps sound right to me. And your idea of integrating with DefinitelyTyped would be a massive step towards adoption (and maybe the only step needed to satisfy reconsideration point 1).

If you go with the ESLint approach, it looks like eslint-plugin-jsdoc already has the first bullet point (via requires-throw). And can be a good source of inspiration.

As far as the spec of how exactly it should work

1. It might be valuable to draft an RFC that covers how it will work (kind of like how it's outlined in the md) - if you're interested in input.
2. It might be easier to forego having a "safe" throws in the first version (if you're looking to shave off scope). And have a stance on checked vs unchecked exceptions.
   • Swift enforces checked exceptions (you have to specify that you throw and you always have to catch or declare throw)
   • Kotlin has no checked exceptions (you can throw, but there's no enforcement you catch - like TypeScript)
     • They support @throws annotation for interop
   • Java has checked exceptions (with some special unchecked ones)

Java seems closest to the hybrid of supporting both (safe and unsafe). But I'd personally advocate for matching swift (checked exceptions), but with the ability to opt into

• None, warning, and error versions of enforcing you declare @throws
• None, warning, and error versions of enforcing you catch something that declares @throws

[KashubaK]

Thanks for your input @kvenn!

The reason I wanted to introduce "safe" exceptions is to address an important note mentioned earlier:

Constructing a RegExp from non-hardcoded input is so vanishingly rare that it seems obnoxious to force a try/catch around new RegExp("[A-Z]") since that technically "can" throw an unavoidable exception, even though by inspection it clearly doesn't.

But perhaps it would be more dangerous and confusing than helpful. Some configuration to whitelist certain functions might be appropriate instead.

Regarding an RFC, for something like this I wouldn't hesitate, but I have some ideas that I need to make sure are even feasible before I potentially waste others' time. Regardless I would love a community effort, as it will be required to move the needle in any meaningful direction.

[KashubaK]

I think it’s important enough to not be closed and encourage more discussion

Personally I agree. That being said, as @kvenn stated, the TS team has already made their stance clear. Robust error handling needs to be normalized in the community before such an invasive new concept is added to countless others' workflows. If our ideas aren't able to proliferate, well, perhaps they were doomed to begin with.

[obedm503]

Constructing a RegExp from non-hardcoded input is so vanishingly rare that it seems obnoxious to force a try/catch

It would be simpler to just use // @ts-expect-error than introducing a whole new error handling concept.

[zefir-git]

Constructing a RegExp from non-hardcoded input is so vanishingly rare that it seems obnoxious to force a try/catch

It would be simpler to just use // @ts-expect-error than introducing a whole new error handling concept.

How would you specify the error type?

[obedm503]

// @ts-expect-error would allow you to leave "safe" errors unhandled.

[KashubaK]

// @ts-expect-error would allow you to leave "safe" errors unhandled.

Yeah, but then you lose all other type-checking for that line.

[michaelangeloio]

I made an extension similar to this!

Feel free to check it out:
https://marketplace.visualstudio.com/items?itemName=michaelangeloio.does-it-throw-vscode

https://github.com/michaelangeloio/does-it-throw

[kvenn]

Very cool! Any plans to build for IntelliJ?

[KashubaK]

@michaelangeloio Super cool, thanks for sharing! I would also love to see a JetBrains plugin implemented.

Any plans/ideas about a CLI? Could it report the class of error? i.e. class MyCustomError extends Error {}

[RyanCavanaugh]

Locking for clarity. This issue is concluded and will not be reopened; see #13219 (comment) . For general discussion on the topic, continue if needed at #56365.