Related
I know ECMAScript 6 has constructors but is there such a thing as destructors for ECMAScript 6?
For example if I register some of my object's methods as event listeners in the constructor, I want to remove them when my object is deleted.
One solution is to have a convention of creating a destructor method for every class that needs this kind of behaviour and manually call it. This will remove the references to the event handlers, hence my object will truly be ready for garbage collection. Otherwise it'll stay in memory because of those methods.
But I was hoping if ECMAScript 6 has something native that will be called right before the object is garbage collected.
If there is no such mechanism, what is a pattern/convention for such problems?
Is there such a thing as destructors for ECMAScript 6?
No. EcmaScript 6 does not specify any garbage collection semantics at all[1], so there is nothing like a "destruction" either.
If I register some of my object's methods as event listeners in the constructor, I want to remove them when my object is deleted
A destructor wouldn't even help you here. It's the event listeners themselves that still reference your object, so it would not be able to get garbage-collected before they are unregistered.
What you are actually looking for is a method of registering listeners without marking them as live root objects. (Ask your local eventsource manufacturer for such a feature).
1): Well, there is a beginning with the specification of WeakMap and WeakSet objects. However, true weak references are still in the pipeline [1][2].
I just came across this question in a search about destructors and I thought there was an unanswered part of your question in your comments, so I thought I would address that.
thank you guys. But what would be a good convention if ECMAScript
doesn't have destructors? Should I create a method called destructor
and call it manually when I'm done with the object? Any other idea?
If you want to tell your object that you are now done with it and it should specifically release any event listeners it has, then you can just create an ordinary method for doing that. You can call the method something like release() or deregister() or unhook() or anything of that ilk. The idea is that you're telling the object to disconnect itself from anything else it is hooked up to (deregister event listeners, clear external object references, etc...). You will have to call it manually at the appropriate time.
If, at the same time you also make sure there are no other references to that object, then your object will become eligible for garbage collection at that point.
ES6 does have weakMap and weakSet which are ways of keeping track of a set of objects that are still alive without affecting when they can be garbage collected, but it does not provide any sort of notification when they are garbage collected. They just disappear from the weakMap or weakSet at some point (when they are GCed).
FYI, the issue with this type of destructor you ask for (and probably why there isn't much of a call for it) is that because of garbage collection, an item is not eligible for garbage collection when it has an open event handler against a live object so even if there was such a destructor, it would never get called in your circumstance until you actually removed the event listeners. And, once you've removed the event listeners, there's no need for the destructor for this purpose.
I suppose there's a possible weakListener() that would not prevent garbage collection, but such a thing does not exist either.
FYI, here's another relevant question Why is the object destructor paradigm in garbage collected languages pervasively absent?. This discussion covers finalizer, destructor and disposer design patterns. I found it useful to see the distinction between the three.
Edit in 2020 - proposal for object finalizer
There is a Stage 3 EMCAScript proposal to add a user-defined finalizer function after an object is garbage collected.
A canonical example of something that would benefit from a feature like this is an object that contains a handle to an open file. If the object is garbage collected (because no other code still has a reference to it), then this finalizer scheme allows one to at least put a message to the console that an external resource has just been leaked and code elsewhere should be fixed to prevent this leak.
If you read the proposal thoroughly, you will see that it's nothing like a full-blown destructor in a language like C++. This finalizer is called after the object has already been destroyed and you have to predetermine what part of the instance data needs to be passed to the finalizer for it to do its work. Further, this feature is not meant to be relied upon for normal operation, but rather as a debugging aid and as a backstop against certain types of bugs. You can read the full explanation for these limitations in the proposal.
You have to manually "destruct" objects in JS. Creating a destroy function is common in JS. In other languages this might be called free, release, dispose, close, etc. In my experience though it tends to be destroy which will unhook internal references, events and possibly propagates destroy calls to child objects as well.
WeakMaps are largely useless as they cannot be iterated and this probably wont be available until ECMA 7 if at all. All WeakMaps let you do is have invisible properties detached from the object itself except for lookup by the object reference and GC so that they don't disturb it. This can be useful for caching, extending and dealing with plurality but it doesn't really help with memory management for observables and observers. WeakSet is a subset of WeakMap (like a WeakMap with a default value of boolean true).
There are various arguments on whether to use various implementations of weak references for this or destructors. Both have potential problems and destructors are more limited.
Destructors are actually potentially useless for observers/listeners as well because typically the listener will hold references to the observer either directly or indirectly. A destructor only really works in a proxy fashion without weak references. If your Observer is really just a proxy taking something else's Listeners and putting them on an observable then it can do something there but this sort of thing is rarely useful. Destructors are more for IO related things or doing things outside of the scope of containment (IE, linking up two instances that it created).
The specific case that I started looking into this for is because I have class A instance that takes class B in the constructor, then creates class C instance which listens to B. I always keep the B instance around somewhere high above. A I sometimes throw away, create new ones, create many, etc. In this situation a Destructor would actually work for me but with a nasty side effect that in the parent if I passed the C instance around but removed all A references then the C and B binding would be broken (C has the ground removed from beneath it).
In JS having no automatic solution is painful but I don't think it's easily solvable. Consider these classes (pseudo):
function Filter(stream) {
stream.on('data', function() {
this.emit('data', data.toString().replace('somenoise', '')); // Pretend chunks/multibyte are not a problem.
});
}
Filter.prototype.__proto__ = EventEmitter.prototype;
function View(df, stream) {
df.on('data', function(data) {
stream.write(data.toUpper()); // Shout.
});
}
On a side note, it's hard to make things work without anonymous/unique functions which will be covered later.
In a normal case instantiation would be as so (pseudo):
var df = new Filter(stdin),
v1 = new View(df, stdout),
v2 = new View(df, stderr);
To GC these normally you would set them to null but it wont work because they've created a tree with stdin at the root. This is basically what event systems do. You give a parent to a child, the child adds itself to the parent and then may or may not maintain a reference to the parent. A tree is a simple example but in reality you may also find yourself with complex graphs albeit rarely.
In this case, Filter adds a reference to itself to stdin in the form of an anonymous function which indirectly references Filter by scope. Scope references are something to be aware of and that can be quite complex. A powerful GC can do some interesting things to carve away at items in scope variables but that's another topic. What is critical to understand is that when you create an anonymous function and add it to something as a listener to ab observable, the observable will maintain a reference to the function and anything the function references in the scopes above it (that it was defined in) will also be maintained. The views do the same but after the execution of their constructors the children do not maintain a reference to their parents.
If I set any or all of the vars declared above to null it isn't going to make a difference to anything (similarly when it finished that "main" scope). They will still be active and pipe data from stdin to stdout and stderr.
If I set them all to null it would be impossible to have them removed or GCed without clearing out the events on stdin or setting stdin to null (assuming it can be freed like this). You basically have a memory leak that way with in effect orphaned objects if the rest of the code needs stdin and has other important events on it prohibiting you from doing the aforementioned.
To get rid of df, v1 and v2 I need to call a destroy method on each of them. In terms of implementation this means that both the Filter and View methods need to keep the reference to the anonymous listener function they create as well as the observable and pass that to removeListener.
On a side note, alternatively you can have an obserable that returns an index to keep track of listeners so that you can add prototyped functions which at least to my understanding should be much better on performance and memory. You still have to keep track of the returned identifier though and pass your object to ensure that the listener is bound to it when called.
A destroy function adds several pains. First is that I would have to call it and free the reference:
df.destroy();
v1.destroy();
v2.destroy();
df = v1 = v2 = null;
This is a minor annoyance as it's a bit more code but that is not the real problem. When I hand these references around to many objects. In this case when exactly do you call destroy? You cannot simply hand these off to other objects. You'll end up with chains of destroys and manual implementation of tracking either through program flow or some other means. You can't fire and forget.
An example of this kind of problem is if I decide that View will also call destroy on df when it is destroyed. If v2 is still around destroying df will break it so destroy cannot simply be relayed to df. Instead when v1 takes df to use it, it would need to then tell df it is used which would raise some counter or similar to df. df's destroy function would decrease than counter and only actually destroy if it is 0. This sort of thing adds a lot of complexity and adds a lot that can go wrong the most obvious of which is destroying something while there is still a reference around somewhere that will be used and circular references (at this point it's no longer a case of managing a counter but a map of referencing objects). When you're thinking of implementing your own reference counters, MM and so on in JS then it's probably deficient.
If WeakSets were iterable, this could be used:
function Observable() {
this.events = {open: new WeakSet(), close: new WeakSet()};
}
Observable.prototype.on = function(type, f) {
this.events[type].add(f);
};
Observable.prototype.emit = function(type, ...args) {
this.events[type].forEach(f => f(...args));
};
Observable.prototype.off = function(type, f) {
this.events[type].delete(f);
};
In this case the owning class must also keep a token reference to f otherwise it will go poof.
If Observable were used instead of EventListener then memory management would be automatic in regards to the event listeners.
Instead of calling destroy on each object this would be enough to fully remove them:
df = v1 = v2 = null;
If you didn't set df to null it would still exist but v1 and v2 would automatically be unhooked.
There are two problems with this approach however.
Problem one is that it adds a new complexity. Sometimes people do not actually want this behaviour. I could create a very large chain of objects linked to each other by events rather than containment (references in constructor scopes or object properties). Eventually a tree and I would only have to pass around the root and worry about that. Freeing the root would conveniently free the entire thing. Both behaviours depending on coding style, etc are useful and when creating reusable objects it's going to be hard to either know what people want, what they have done, what you have done and a pain to work around what has been done. If I use Observable instead of EventListener then either df will need to reference v1 and v2 or I'll have to pass them all if I want to transfer ownership of the reference to something else out of scope. A weak reference like thing would mitigate the problem a little by transferring control from Observable to an observer but would not solve it entirely (and needs check on every emit or event on itself). This problem can be fixed I suppose if the behaviour only applies to isolated graphs which would complicate the GC severely and would not apply to cases where there are references outside the graph that are in practice noops (only consume CPU cycles, no changes made).
Problem two is that either it is unpredictable in certain cases or forces the JS engine to traverse the GC graph for those objects on demand which can have a horrific performance impact (although if it is clever it can avoid doing it per member by doing it per WeakMap loop instead). The GC may never run if memory usage does not reach a certain threshold and the object with its events wont be removed. If I set v1 to null it may still relay to stdout forever. Even if it does get GCed this will be arbitrary, it may continue to relay to stdout for any amount of time (1 lines, 10 lines, 2.5 lines, etc).
The reason WeakMap gets away with not caring about the GC when non-iterable is that to access an object you have to have a reference to it anyway so either it hasn't been GCed or hasn't been added to the map.
I am not sure what I think about this kind of thing. You're sort of breaking memory management to fix it with the iterable WeakMap approach. Problem two can also exist for destructors as well.
All of this invokes several levels of hell so I would suggest to try to work around it with good program design, good practices, avoiding certain things, etc. It can be frustrating in JS however because of how flexible it is in certain aspects and because it is more naturally asynchronous and event based with heavy inversion of control.
There is one other solution that is fairly elegant but again still has some potentially serious hangups. If you have a class that extends an observable class you can override the event functions. Add your events to other observables only when events are added to yourself. When all events are removed from you then remove your events from children. You can also make a class to extend your observable class to do this for you. Such a class could provide hooks for empty and non-empty so in a since you would be Observing yourself. This approach isn't bad but also has hangups. There is a complexity increase as well as performance decrease. You'll have to keep a reference to object you observe. Critically, it also will not work for leaves but at least the intermediates will self destruct if you destroy the leaf. It's like chaining destroy but hidden behind calls that you already have to chain. A large performance problem is with this however is that you may have to reinitialise internal data from the Observable everytime your class becomes active. If this process takes a very long time then you might be in trouble.
If you could iterate WeakMap then you could perhaps combine things (switch to Weak when no events, Strong when events) but all that is really doing is putting the performance problem on someone else.
There are also immediate annoyances with iterable WeakMap when it comes to behaviour. I mentioned briefly before about functions having scope references and carving. If I instantiate a child that in the constructor that hooks the listener 'console.log(param)' to parent and fails to persist the parent then when I remove all references to the child it could be freed entirely as the anonymous function added to the parent references nothing from within the child. This leaves the question of what to do about parent.weakmap.add(child, (param) => console.log(param)). To my knowledge the key is weak but not the value so weakmap.add(object, object) is persistent. This is something I need to reevaluate though. To me that looks like a memory leak if I dispose all other object references but I suspect in reality it manages that basically by seeing it as a circular reference. Either the anonymous function maintains an implicit reference to objects resulting from parent scopes for consistency wasting a lot of memory or you have behaviour varying based on circumstances which is hard to predict or manage. I think the former is actually impossible. In the latter case if I have a method on a class that simply takes an object and adds console.log it would be freed when I clear the references to the class even if I returned the function and maintained a reference. To be fair this particular scenario is rarely needed legitimately but eventually someone will find an angle and will be asking for a HalfWeakMap which is iterable (free on key and value refs released) but that is unpredictable as well (obj = null magically ending IO, f = null magically ending IO, both doable at incredible distances).
If there is no such mechanism, what is a pattern/convention for such problems?
The term 'cleanup' might be more appropriate, but will use 'destructor' to match OP
Suppose you write some javascript entirely with 'function's and 'var's.
Then you can use the pattern of writing all the functions code within the framework of a try/catch/finally lattice. Within finally perform the destruction code.
Instead of the C++ style of writing object classes with unspecified lifetimes, and then specifying the lifetime by arbitrary scopes and the implicit call to ~() at scope end (~() is destructor in C++), in this javascript pattern the object is the function, the scope is exactly the function scope, and the destructor is the finally block.
If you are now thinking this pattern is inherently flawed because try/catch/finally doesn't encompass asynchronous execution which is essential to javascript, then you are correct. Fortunately, since 2018 the asynchronous programming helper object Promise has had a prototype function finally added to the already existing resolve and catch prototype functions. That means that that asynchronous scopes requiring destructors can be written with a Promise object, using finally as the destructor. Furthermore you can use try/catch/finally in an async function calling Promises with or without await, but must be aware that Promises called without await will be execute asynchronously outside the scope and so handle the desctructor code in a final then.
In the following code PromiseA and PromiseB are some legacy API level promises which don't have finally function arguments specified. PromiseC DOES have a finally argument defined.
async function afunc(a,b){
try {
function resolveB(r){ ... }
function catchB(e){ ... }
function cleanupB(){ ... }
function resolveC(r){ ... }
function catchC(e){ ... }
function cleanupC(){ ... }
...
// PromiseA preced by await sp will finish before finally block.
// If no rush then safe to handle PromiseA cleanup in finally block
var x = await PromiseA(a);
// PromiseB,PromiseC not preceded by await - will execute asynchronously
// so might finish after finally block so we must provide
// explicit cleanup (if necessary)
PromiseB(b).then(resolveB,catchB).then(cleanupB,cleanupB);
PromiseC(c).then(resolveC,catchC,cleanupC);
}
catch(e) { ... }
finally { /* scope destructor/cleanup code here */ }
}
I am not advocating that every object in javascript be written as a function. Instead, consider the case where you have a scope identified which really 'wants' a destructor to be called at its end of life. Formulate that scope as a function object, using the pattern's finally block (or finally function in the case of an asynchronous scope) as the destructor. It is quite like likely that formulating that functional object obviated the need for a non-function class which would otherwise have been written - no extra code was required, aligning scope and class might even be cleaner.
Note: As others have written, we should not confuse destructors and garbage collection. As it happens C++ destructors are often or mainly concerned with manual garbage collection, but not exclusively so. Javascript has no need for manual garbage collection, but asynchronous scope end-of-life is often a place for (de)registering event listeners, etc..
Here you go. The Subscribe/Publish object will unsubscribe a callback function automatically if it goes out of scope and gets garbage collected.
const createWeakPublisher = () => {
const weakSet = new WeakSet();
const subscriptions = new Set();
return {
subscribe(callback) {
if (!weakSet.has(callback)) {
weakSet.add(callback);
subscriptions.add(new WeakRef(callback));
}
return callback;
},
publish() {
for (const weakRef of subscriptions) {
const callback = weakRef.deref();
console.log(callback?.toString());
if (callback) callback();
else subscriptions.delete(weakRef);
}
},
};
};
Although it might not happen immediately after the callback function goes out of scope, or it might not happen at all. See weakRef documentation for more details. But it works like a charm for my use case.
You might also want to check out the FinalizationRegistry API for a different approach.
"A destructor wouldn't even help you here. It's the event listeners
themselves that still reference your object, so it would not be able
to get garbage-collected before they are unregistered."
Not so. The purpose of a destructor is to allow the item that registered the listeners to unregister them. Once an object has no other references to it, it will be garbage collected.
For instance, in AngularJS, when a controller is destroyed, it can listen for a destroy event and respond to it. This isn't the same as having a destructor automatically called, but it's close, and gives us the opportunity to remove listeners that were set when the controller was initialized.
// Set event listeners, hanging onto the returned listener removal functions
function initialize() {
$scope.listenerCleanup = [];
$scope.listenerCleanup.push( $scope.$on( EVENTS.DESTROY, instance.onDestroy) );
$scope.listenerCleanup.push( $scope.$on( AUTH_SERVICE_RESPONSES.CREATE_USER.SUCCESS, instance.onCreateUserResponse ) );
$scope.listenerCleanup.push( $scope.$on( AUTH_SERVICE_RESPONSES.CREATE_USER.FAILURE, instance.onCreateUserResponse ) );
}
// Remove event listeners when the controller is destroyed
function onDestroy(){
$scope.listenerCleanup.forEach( remove => remove() );
}
Javascript does not have destructures the same way C++ does. Instead, alternative design patterns should be used to manage resources. Here are a couple of examples:
You can restrict users to using the instance for the duration of a callback, after which it'll automatically be cleaned up. (This pattern is similar to the beloved "with" statement in Python)
connectToDatabase(async db => {
const resource = await db.doSomeRequest()
await useResource(resource)
}) // The db connection is closed once the callback ends
When the above example is too restrictive, another alternative is to just create explicit cleanup functions.
const db = makeDatabaseConnection()
const resource = await db.doSomeRequest()
updatePageWithResource(resource)
pageChangeEvent.addListener(() => {
db.destroy()
})
The other answers already explained in detail that there is no destructor. But your actual goal seems to be event related. You have an object which is connected to some event and you want this connection to go away automatically when the object is garbage collected. But this won't happen because the event subscription itself references the listener function. Well, UNLESS you use this nifty new WeakRef stuff.
Here is an example:
<!DOCTYPE html>
<html>
<body>
<button onclick="subscribe()">Subscribe</button>
<button id="emitter">Emit</button>
<button onclick="free()">Free</button>
<script>
const emitter = document.getElementById("emitter");
let listener = null;
function addWeakEventListener(element, event, callback) {
// Weakrefs only can store objects, so we put the callback into an object
const weakRef = new WeakRef({ callback });
const listener = () => {
const obj = weakRef.deref();
if (obj == null) {
console.log("Removing garbage collected event listener");
element.removeEventListener(event, listener);
} else {
obj.callback();
}
};
element.addEventListener(event, listener);
}
function subscribe() {
listener = () => console.log("Event fired!");
addWeakEventListener(emitter, "click", listener);
console.log("Listener created and subscribed to emitter");
}
function free() {
listener = null;
console.log("Reference cleared. Now force garbage collection in dev console or wait some time before clicking Emit again.");
}
</script>
</body>
</html>
(JSFiddle)
Clicking the Subscribe button creates a new listener function and registers it at the click event of the Emit button. So clicking the Emit button after that prints a message to the console. Now click the Free button which simply sets the listener variable to null so the garbage collector can remove the listener. Wait some time or force gargabe collection in the developer console and then click the Emit button again. The wrapper listener function now sees that the actual listener (wrapped in a WeakRef) is no longer there and then unsubscribes itself from the button.
WeakRefs are quite powerful but note that there is no guarantee if and when your stuff is garbage collected.
The answer to the question as-stated in the title is FinalizationRegistry, available since Firefox 79 (June 2020), Chrome 84 and derivatives (July 2020), Safari 14.1 (April 2021), and Node 14.6.0 (July 2020)… however, a native JS destructor is probably not the right solution for your use-case.
function create_eval_worker(f) {
let src_worker_blob = new Blob([f.toString()], {type: 'application/javascript'});
let src_worker_url = URL.createObjectURL(src_worker_blob);
async function g() {
let w = new Worker(src_worker_url);
…
}
// Run URL.revokeObjectURL(src_worker_url) as a destructor of g
let registry = new FinalizationRegistry(u => URL.revokeObjectURL(u));
registry.register(g, src_worker_url);
return g;
}
}
Caveat:
Avoid where possible
Correct use of FinalizationRegistry takes careful thought, and it's best avoided if possible. When, how, and whether garbage collection occurs is down to the implementation of any given JavaScript engine. Any behavior you observe in one engine may be different in another engine, in another version of the same engine, or even in a slightly different situation with the same version of the same engine.
…
Developers shouldn't rely on cleanup callbacks for essential program logic. Cleanup callbacks may be useful for reducing memory usage across the course of a program, but are unlikely to be useful otherwise.
A conforming JavaScript implementation, even one that does garbage collection, is not required to call cleanup callbacks. When and whether it does so is entirely down to the implementation of the JavaScript engine. When a registered object is reclaimed, any cleanup callbacks for it may be called then, or some time later, or not at all.
–Mozilla Developer Network
I know ECMAScript 6 has constructors but is there such a thing as destructors for ECMAScript 6?
For example if I register some of my object's methods as event listeners in the constructor, I want to remove them when my object is deleted.
One solution is to have a convention of creating a destructor method for every class that needs this kind of behaviour and manually call it. This will remove the references to the event handlers, hence my object will truly be ready for garbage collection. Otherwise it'll stay in memory because of those methods.
But I was hoping if ECMAScript 6 has something native that will be called right before the object is garbage collected.
If there is no such mechanism, what is a pattern/convention for such problems?
Is there such a thing as destructors for ECMAScript 6?
No. EcmaScript 6 does not specify any garbage collection semantics at all[1], so there is nothing like a "destruction" either.
If I register some of my object's methods as event listeners in the constructor, I want to remove them when my object is deleted
A destructor wouldn't even help you here. It's the event listeners themselves that still reference your object, so it would not be able to get garbage-collected before they are unregistered.
What you are actually looking for is a method of registering listeners without marking them as live root objects. (Ask your local eventsource manufacturer for such a feature).
1): Well, there is a beginning with the specification of WeakMap and WeakSet objects. However, true weak references are still in the pipeline [1][2].
I just came across this question in a search about destructors and I thought there was an unanswered part of your question in your comments, so I thought I would address that.
thank you guys. But what would be a good convention if ECMAScript
doesn't have destructors? Should I create a method called destructor
and call it manually when I'm done with the object? Any other idea?
If you want to tell your object that you are now done with it and it should specifically release any event listeners it has, then you can just create an ordinary method for doing that. You can call the method something like release() or deregister() or unhook() or anything of that ilk. The idea is that you're telling the object to disconnect itself from anything else it is hooked up to (deregister event listeners, clear external object references, etc...). You will have to call it manually at the appropriate time.
If, at the same time you also make sure there are no other references to that object, then your object will become eligible for garbage collection at that point.
ES6 does have weakMap and weakSet which are ways of keeping track of a set of objects that are still alive without affecting when they can be garbage collected, but it does not provide any sort of notification when they are garbage collected. They just disappear from the weakMap or weakSet at some point (when they are GCed).
FYI, the issue with this type of destructor you ask for (and probably why there isn't much of a call for it) is that because of garbage collection, an item is not eligible for garbage collection when it has an open event handler against a live object so even if there was such a destructor, it would never get called in your circumstance until you actually removed the event listeners. And, once you've removed the event listeners, there's no need for the destructor for this purpose.
I suppose there's a possible weakListener() that would not prevent garbage collection, but such a thing does not exist either.
FYI, here's another relevant question Why is the object destructor paradigm in garbage collected languages pervasively absent?. This discussion covers finalizer, destructor and disposer design patterns. I found it useful to see the distinction between the three.
Edit in 2020 - proposal for object finalizer
There is a Stage 3 EMCAScript proposal to add a user-defined finalizer function after an object is garbage collected.
A canonical example of something that would benefit from a feature like this is an object that contains a handle to an open file. If the object is garbage collected (because no other code still has a reference to it), then this finalizer scheme allows one to at least put a message to the console that an external resource has just been leaked and code elsewhere should be fixed to prevent this leak.
If you read the proposal thoroughly, you will see that it's nothing like a full-blown destructor in a language like C++. This finalizer is called after the object has already been destroyed and you have to predetermine what part of the instance data needs to be passed to the finalizer for it to do its work. Further, this feature is not meant to be relied upon for normal operation, but rather as a debugging aid and as a backstop against certain types of bugs. You can read the full explanation for these limitations in the proposal.
You have to manually "destruct" objects in JS. Creating a destroy function is common in JS. In other languages this might be called free, release, dispose, close, etc. In my experience though it tends to be destroy which will unhook internal references, events and possibly propagates destroy calls to child objects as well.
WeakMaps are largely useless as they cannot be iterated and this probably wont be available until ECMA 7 if at all. All WeakMaps let you do is have invisible properties detached from the object itself except for lookup by the object reference and GC so that they don't disturb it. This can be useful for caching, extending and dealing with plurality but it doesn't really help with memory management for observables and observers. WeakSet is a subset of WeakMap (like a WeakMap with a default value of boolean true).
There are various arguments on whether to use various implementations of weak references for this or destructors. Both have potential problems and destructors are more limited.
Destructors are actually potentially useless for observers/listeners as well because typically the listener will hold references to the observer either directly or indirectly. A destructor only really works in a proxy fashion without weak references. If your Observer is really just a proxy taking something else's Listeners and putting them on an observable then it can do something there but this sort of thing is rarely useful. Destructors are more for IO related things or doing things outside of the scope of containment (IE, linking up two instances that it created).
The specific case that I started looking into this for is because I have class A instance that takes class B in the constructor, then creates class C instance which listens to B. I always keep the B instance around somewhere high above. A I sometimes throw away, create new ones, create many, etc. In this situation a Destructor would actually work for me but with a nasty side effect that in the parent if I passed the C instance around but removed all A references then the C and B binding would be broken (C has the ground removed from beneath it).
In JS having no automatic solution is painful but I don't think it's easily solvable. Consider these classes (pseudo):
function Filter(stream) {
stream.on('data', function() {
this.emit('data', data.toString().replace('somenoise', '')); // Pretend chunks/multibyte are not a problem.
});
}
Filter.prototype.__proto__ = EventEmitter.prototype;
function View(df, stream) {
df.on('data', function(data) {
stream.write(data.toUpper()); // Shout.
});
}
On a side note, it's hard to make things work without anonymous/unique functions which will be covered later.
In a normal case instantiation would be as so (pseudo):
var df = new Filter(stdin),
v1 = new View(df, stdout),
v2 = new View(df, stderr);
To GC these normally you would set them to null but it wont work because they've created a tree with stdin at the root. This is basically what event systems do. You give a parent to a child, the child adds itself to the parent and then may or may not maintain a reference to the parent. A tree is a simple example but in reality you may also find yourself with complex graphs albeit rarely.
In this case, Filter adds a reference to itself to stdin in the form of an anonymous function which indirectly references Filter by scope. Scope references are something to be aware of and that can be quite complex. A powerful GC can do some interesting things to carve away at items in scope variables but that's another topic. What is critical to understand is that when you create an anonymous function and add it to something as a listener to ab observable, the observable will maintain a reference to the function and anything the function references in the scopes above it (that it was defined in) will also be maintained. The views do the same but after the execution of their constructors the children do not maintain a reference to their parents.
If I set any or all of the vars declared above to null it isn't going to make a difference to anything (similarly when it finished that "main" scope). They will still be active and pipe data from stdin to stdout and stderr.
If I set them all to null it would be impossible to have them removed or GCed without clearing out the events on stdin or setting stdin to null (assuming it can be freed like this). You basically have a memory leak that way with in effect orphaned objects if the rest of the code needs stdin and has other important events on it prohibiting you from doing the aforementioned.
To get rid of df, v1 and v2 I need to call a destroy method on each of them. In terms of implementation this means that both the Filter and View methods need to keep the reference to the anonymous listener function they create as well as the observable and pass that to removeListener.
On a side note, alternatively you can have an obserable that returns an index to keep track of listeners so that you can add prototyped functions which at least to my understanding should be much better on performance and memory. You still have to keep track of the returned identifier though and pass your object to ensure that the listener is bound to it when called.
A destroy function adds several pains. First is that I would have to call it and free the reference:
df.destroy();
v1.destroy();
v2.destroy();
df = v1 = v2 = null;
This is a minor annoyance as it's a bit more code but that is not the real problem. When I hand these references around to many objects. In this case when exactly do you call destroy? You cannot simply hand these off to other objects. You'll end up with chains of destroys and manual implementation of tracking either through program flow or some other means. You can't fire and forget.
An example of this kind of problem is if I decide that View will also call destroy on df when it is destroyed. If v2 is still around destroying df will break it so destroy cannot simply be relayed to df. Instead when v1 takes df to use it, it would need to then tell df it is used which would raise some counter or similar to df. df's destroy function would decrease than counter and only actually destroy if it is 0. This sort of thing adds a lot of complexity and adds a lot that can go wrong the most obvious of which is destroying something while there is still a reference around somewhere that will be used and circular references (at this point it's no longer a case of managing a counter but a map of referencing objects). When you're thinking of implementing your own reference counters, MM and so on in JS then it's probably deficient.
If WeakSets were iterable, this could be used:
function Observable() {
this.events = {open: new WeakSet(), close: new WeakSet()};
}
Observable.prototype.on = function(type, f) {
this.events[type].add(f);
};
Observable.prototype.emit = function(type, ...args) {
this.events[type].forEach(f => f(...args));
};
Observable.prototype.off = function(type, f) {
this.events[type].delete(f);
};
In this case the owning class must also keep a token reference to f otherwise it will go poof.
If Observable were used instead of EventListener then memory management would be automatic in regards to the event listeners.
Instead of calling destroy on each object this would be enough to fully remove them:
df = v1 = v2 = null;
If you didn't set df to null it would still exist but v1 and v2 would automatically be unhooked.
There are two problems with this approach however.
Problem one is that it adds a new complexity. Sometimes people do not actually want this behaviour. I could create a very large chain of objects linked to each other by events rather than containment (references in constructor scopes or object properties). Eventually a tree and I would only have to pass around the root and worry about that. Freeing the root would conveniently free the entire thing. Both behaviours depending on coding style, etc are useful and when creating reusable objects it's going to be hard to either know what people want, what they have done, what you have done and a pain to work around what has been done. If I use Observable instead of EventListener then either df will need to reference v1 and v2 or I'll have to pass them all if I want to transfer ownership of the reference to something else out of scope. A weak reference like thing would mitigate the problem a little by transferring control from Observable to an observer but would not solve it entirely (and needs check on every emit or event on itself). This problem can be fixed I suppose if the behaviour only applies to isolated graphs which would complicate the GC severely and would not apply to cases where there are references outside the graph that are in practice noops (only consume CPU cycles, no changes made).
Problem two is that either it is unpredictable in certain cases or forces the JS engine to traverse the GC graph for those objects on demand which can have a horrific performance impact (although if it is clever it can avoid doing it per member by doing it per WeakMap loop instead). The GC may never run if memory usage does not reach a certain threshold and the object with its events wont be removed. If I set v1 to null it may still relay to stdout forever. Even if it does get GCed this will be arbitrary, it may continue to relay to stdout for any amount of time (1 lines, 10 lines, 2.5 lines, etc).
The reason WeakMap gets away with not caring about the GC when non-iterable is that to access an object you have to have a reference to it anyway so either it hasn't been GCed or hasn't been added to the map.
I am not sure what I think about this kind of thing. You're sort of breaking memory management to fix it with the iterable WeakMap approach. Problem two can also exist for destructors as well.
All of this invokes several levels of hell so I would suggest to try to work around it with good program design, good practices, avoiding certain things, etc. It can be frustrating in JS however because of how flexible it is in certain aspects and because it is more naturally asynchronous and event based with heavy inversion of control.
There is one other solution that is fairly elegant but again still has some potentially serious hangups. If you have a class that extends an observable class you can override the event functions. Add your events to other observables only when events are added to yourself. When all events are removed from you then remove your events from children. You can also make a class to extend your observable class to do this for you. Such a class could provide hooks for empty and non-empty so in a since you would be Observing yourself. This approach isn't bad but also has hangups. There is a complexity increase as well as performance decrease. You'll have to keep a reference to object you observe. Critically, it also will not work for leaves but at least the intermediates will self destruct if you destroy the leaf. It's like chaining destroy but hidden behind calls that you already have to chain. A large performance problem is with this however is that you may have to reinitialise internal data from the Observable everytime your class becomes active. If this process takes a very long time then you might be in trouble.
If you could iterate WeakMap then you could perhaps combine things (switch to Weak when no events, Strong when events) but all that is really doing is putting the performance problem on someone else.
There are also immediate annoyances with iterable WeakMap when it comes to behaviour. I mentioned briefly before about functions having scope references and carving. If I instantiate a child that in the constructor that hooks the listener 'console.log(param)' to parent and fails to persist the parent then when I remove all references to the child it could be freed entirely as the anonymous function added to the parent references nothing from within the child. This leaves the question of what to do about parent.weakmap.add(child, (param) => console.log(param)). To my knowledge the key is weak but not the value so weakmap.add(object, object) is persistent. This is something I need to reevaluate though. To me that looks like a memory leak if I dispose all other object references but I suspect in reality it manages that basically by seeing it as a circular reference. Either the anonymous function maintains an implicit reference to objects resulting from parent scopes for consistency wasting a lot of memory or you have behaviour varying based on circumstances which is hard to predict or manage. I think the former is actually impossible. In the latter case if I have a method on a class that simply takes an object and adds console.log it would be freed when I clear the references to the class even if I returned the function and maintained a reference. To be fair this particular scenario is rarely needed legitimately but eventually someone will find an angle and will be asking for a HalfWeakMap which is iterable (free on key and value refs released) but that is unpredictable as well (obj = null magically ending IO, f = null magically ending IO, both doable at incredible distances).
If there is no such mechanism, what is a pattern/convention for such problems?
The term 'cleanup' might be more appropriate, but will use 'destructor' to match OP
Suppose you write some javascript entirely with 'function's and 'var's.
Then you can use the pattern of writing all the functions code within the framework of a try/catch/finally lattice. Within finally perform the destruction code.
Instead of the C++ style of writing object classes with unspecified lifetimes, and then specifying the lifetime by arbitrary scopes and the implicit call to ~() at scope end (~() is destructor in C++), in this javascript pattern the object is the function, the scope is exactly the function scope, and the destructor is the finally block.
If you are now thinking this pattern is inherently flawed because try/catch/finally doesn't encompass asynchronous execution which is essential to javascript, then you are correct. Fortunately, since 2018 the asynchronous programming helper object Promise has had a prototype function finally added to the already existing resolve and catch prototype functions. That means that that asynchronous scopes requiring destructors can be written with a Promise object, using finally as the destructor. Furthermore you can use try/catch/finally in an async function calling Promises with or without await, but must be aware that Promises called without await will be execute asynchronously outside the scope and so handle the desctructor code in a final then.
In the following code PromiseA and PromiseB are some legacy API level promises which don't have finally function arguments specified. PromiseC DOES have a finally argument defined.
async function afunc(a,b){
try {
function resolveB(r){ ... }
function catchB(e){ ... }
function cleanupB(){ ... }
function resolveC(r){ ... }
function catchC(e){ ... }
function cleanupC(){ ... }
...
// PromiseA preced by await sp will finish before finally block.
// If no rush then safe to handle PromiseA cleanup in finally block
var x = await PromiseA(a);
// PromiseB,PromiseC not preceded by await - will execute asynchronously
// so might finish after finally block so we must provide
// explicit cleanup (if necessary)
PromiseB(b).then(resolveB,catchB).then(cleanupB,cleanupB);
PromiseC(c).then(resolveC,catchC,cleanupC);
}
catch(e) { ... }
finally { /* scope destructor/cleanup code here */ }
}
I am not advocating that every object in javascript be written as a function. Instead, consider the case where you have a scope identified which really 'wants' a destructor to be called at its end of life. Formulate that scope as a function object, using the pattern's finally block (or finally function in the case of an asynchronous scope) as the destructor. It is quite like likely that formulating that functional object obviated the need for a non-function class which would otherwise have been written - no extra code was required, aligning scope and class might even be cleaner.
Note: As others have written, we should not confuse destructors and garbage collection. As it happens C++ destructors are often or mainly concerned with manual garbage collection, but not exclusively so. Javascript has no need for manual garbage collection, but asynchronous scope end-of-life is often a place for (de)registering event listeners, etc..
Here you go. The Subscribe/Publish object will unsubscribe a callback function automatically if it goes out of scope and gets garbage collected.
const createWeakPublisher = () => {
const weakSet = new WeakSet();
const subscriptions = new Set();
return {
subscribe(callback) {
if (!weakSet.has(callback)) {
weakSet.add(callback);
subscriptions.add(new WeakRef(callback));
}
return callback;
},
publish() {
for (const weakRef of subscriptions) {
const callback = weakRef.deref();
console.log(callback?.toString());
if (callback) callback();
else subscriptions.delete(weakRef);
}
},
};
};
Although it might not happen immediately after the callback function goes out of scope, or it might not happen at all. See weakRef documentation for more details. But it works like a charm for my use case.
You might also want to check out the FinalizationRegistry API for a different approach.
"A destructor wouldn't even help you here. It's the event listeners
themselves that still reference your object, so it would not be able
to get garbage-collected before they are unregistered."
Not so. The purpose of a destructor is to allow the item that registered the listeners to unregister them. Once an object has no other references to it, it will be garbage collected.
For instance, in AngularJS, when a controller is destroyed, it can listen for a destroy event and respond to it. This isn't the same as having a destructor automatically called, but it's close, and gives us the opportunity to remove listeners that were set when the controller was initialized.
// Set event listeners, hanging onto the returned listener removal functions
function initialize() {
$scope.listenerCleanup = [];
$scope.listenerCleanup.push( $scope.$on( EVENTS.DESTROY, instance.onDestroy) );
$scope.listenerCleanup.push( $scope.$on( AUTH_SERVICE_RESPONSES.CREATE_USER.SUCCESS, instance.onCreateUserResponse ) );
$scope.listenerCleanup.push( $scope.$on( AUTH_SERVICE_RESPONSES.CREATE_USER.FAILURE, instance.onCreateUserResponse ) );
}
// Remove event listeners when the controller is destroyed
function onDestroy(){
$scope.listenerCleanup.forEach( remove => remove() );
}
Javascript does not have destructures the same way C++ does. Instead, alternative design patterns should be used to manage resources. Here are a couple of examples:
You can restrict users to using the instance for the duration of a callback, after which it'll automatically be cleaned up. (This pattern is similar to the beloved "with" statement in Python)
connectToDatabase(async db => {
const resource = await db.doSomeRequest()
await useResource(resource)
}) // The db connection is closed once the callback ends
When the above example is too restrictive, another alternative is to just create explicit cleanup functions.
const db = makeDatabaseConnection()
const resource = await db.doSomeRequest()
updatePageWithResource(resource)
pageChangeEvent.addListener(() => {
db.destroy()
})
The other answers already explained in detail that there is no destructor. But your actual goal seems to be event related. You have an object which is connected to some event and you want this connection to go away automatically when the object is garbage collected. But this won't happen because the event subscription itself references the listener function. Well, UNLESS you use this nifty new WeakRef stuff.
Here is an example:
<!DOCTYPE html>
<html>
<body>
<button onclick="subscribe()">Subscribe</button>
<button id="emitter">Emit</button>
<button onclick="free()">Free</button>
<script>
const emitter = document.getElementById("emitter");
let listener = null;
function addWeakEventListener(element, event, callback) {
// Weakrefs only can store objects, so we put the callback into an object
const weakRef = new WeakRef({ callback });
const listener = () => {
const obj = weakRef.deref();
if (obj == null) {
console.log("Removing garbage collected event listener");
element.removeEventListener(event, listener);
} else {
obj.callback();
}
};
element.addEventListener(event, listener);
}
function subscribe() {
listener = () => console.log("Event fired!");
addWeakEventListener(emitter, "click", listener);
console.log("Listener created and subscribed to emitter");
}
function free() {
listener = null;
console.log("Reference cleared. Now force garbage collection in dev console or wait some time before clicking Emit again.");
}
</script>
</body>
</html>
(JSFiddle)
Clicking the Subscribe button creates a new listener function and registers it at the click event of the Emit button. So clicking the Emit button after that prints a message to the console. Now click the Free button which simply sets the listener variable to null so the garbage collector can remove the listener. Wait some time or force gargabe collection in the developer console and then click the Emit button again. The wrapper listener function now sees that the actual listener (wrapped in a WeakRef) is no longer there and then unsubscribes itself from the button.
WeakRefs are quite powerful but note that there is no guarantee if and when your stuff is garbage collected.
The answer to the question as-stated in the title is FinalizationRegistry, available since Firefox 79 (June 2020), Chrome 84 and derivatives (July 2020), Safari 14.1 (April 2021), and Node 14.6.0 (July 2020)… however, a native JS destructor is probably not the right solution for your use-case.
function create_eval_worker(f) {
let src_worker_blob = new Blob([f.toString()], {type: 'application/javascript'});
let src_worker_url = URL.createObjectURL(src_worker_blob);
async function g() {
let w = new Worker(src_worker_url);
…
}
// Run URL.revokeObjectURL(src_worker_url) as a destructor of g
let registry = new FinalizationRegistry(u => URL.revokeObjectURL(u));
registry.register(g, src_worker_url);
return g;
}
}
Caveat:
Avoid where possible
Correct use of FinalizationRegistry takes careful thought, and it's best avoided if possible. When, how, and whether garbage collection occurs is down to the implementation of any given JavaScript engine. Any behavior you observe in one engine may be different in another engine, in another version of the same engine, or even in a slightly different situation with the same version of the same engine.
…
Developers shouldn't rely on cleanup callbacks for essential program logic. Cleanup callbacks may be useful for reducing memory usage across the course of a program, but are unlikely to be useful otherwise.
A conforming JavaScript implementation, even one that does garbage collection, is not required to call cleanup callbacks. When and whether it does so is entirely down to the implementation of the JavaScript engine. When a registered object is reclaimed, any cleanup callbacks for it may be called then, or some time later, or not at all.
–Mozilla Developer Network
In my multiuser live nods.js app, I have a function that removes a specific element from an array. It's an asyn function and can potentially be called by two different users independently and probably at the same time (client-site socket trigger) to perform the operation on the same array.
myUsers= [];
myUsers.push('user1');
myUsers.push('user2');
myUsers.push('user3');
function removeUser(user){
for(var i=0; i<=myUsers.length; i++){
if(user===myUsers[i]){
myUsers.splice(i,1);
break;
} } }
Should I be worried that it might lead to removing the wrong/unintended element as a result of both operations happening simultaneously? Like, when removeUser('user1') and removeUser('user3') are both called at the same time (client-site socket trigger), and while user1 is being spliced, and user3 has reached at i=2 it might remove user2 instead of user3? And i might even now go over the loop's original limit (myUsers.length)?
Is my concern valid? If so, what would be the solution?
No. Node.js is single threaded. So, only one function will be executed at a time. The problem you describe is specific to multi-threaded environments only.
For example, though removeUser('user1') and removeUser('user3') are ready to be executed, they will be pushed in a queue. And the Node.js engine will take the element at the front and executes it and once finishes executing that, picks the next item from the queue. So, user1 will be removed first and then user3.
I've been told a few times that it's "bad" to use exceptions in javascript. Haven't really been told why it's bad, but inestead that I should use break, continue and return instead.
Which is fine, except I don't need to return/break/continue, I need to throw. I have a case where I have iterator functions nested in each other, each one returns the next value on call, so I use an exception to indicate that there's nothing more to iterate : seems like a logical way to do it, makes the code clean and works perfectly in practice. Is there really any reason not to use exceptions in js?
Second question, when I'm using exceptions, what kind of objects should I throw? For errors I'll obviously throw instances of Error, but for special cases (stop iteration, etc) I needed a quick way to check for those specific exceptions, and what I did is simply define an empty named function (function StopIteration(){}) and since functions are compared by reference I can always check if it's my special case or I should just rethrow. Is there a better or more idomatic way to do this in js? Should I really try to refactor my code avoid using exceptions?
Thank you
It sounds like you're searching through a multidimensional space using nested loops, which is a fairly common thing to program. It's tempting to throw an exception from the innermost loop when the target is found which then gets caught by a "catch" block around the outermost loop but this often considered poor practice.
Language designers are aware of this pitfall so they allow us to give labels (names) to loop structures so that we can break from them (or continue to them). Consider this example:
function findValueInMatrix(value, matrix) {
var r, c, coords, rows=matrix.length, cols=matrix[0].length;
found: // Label the outer loop as "found" so we can break from it.
for (r=0; r<rows; r++) {
for (c=0; c<cols; c++) {
if (matrix[r][c] == value) {
coords = [r, c]
break found; // Exit the loop labeled "found".
}
}
}
return coords;
}
You can find more information in this post about breaking from nested loops.
This jsPerf test case also demonstrates that breaking to a label is nominally faster than throwing an exception (presumably in most browsers).
exception handling is in general slower than normal evaluation and only for exceptional situations
for stopping iterations you are better of defining a hasNext or isEmpty function for an ending condition or return a sentinel value (like undefined) when there are no more elements so the loop becomes
//with hasNext
while(it.hasNext()){
var val = it.next();
//...
}
//or with a sentinal undefined
while( (val = it.next()) !== undefined){
//...
}
For some people, throwing exceptions to indicate that there is no more values is a bad style; exceptions are made for exceptional behaviour.
Moreover, exception handling is generally slower than simple tests.
The difference is really minor (and maybe does not even exist) if the thrown instance is not created everytime. So having a constant StopIteration and throwing it is in fact quite fast.
For your problem, you might want to see https://developer.mozilla.org/en/JavaScript/Guide/Iterators_and_Generators (they use exceptions to stop iteration).
I found that in dynamically typed languages exceptions are used for this kind of purposes too.
Throwing an exception to break out of a nested recursive search/walk is the easiest and more elegant way to go.
Say you walk a tree, recursivelly calling a function to handle the current node, calling again for each of its children and so on.
As you process each node, you might want to completelly break the whole call stack and return the found value (or stop iterating in general), breaking out of all the loops that are waiting in previous calls. Mind you that these previous calls might be costly in processing & memory, but useless if you "found" your result. Throwing an Exception is the most efficient and simple way to achieve this, but of course it needs attention.
For simple (even nested) loops its a overkill and is rightly touted as an anti-pattern.
There is array of objects, which are expanded with parallel ajax requests. When last request is done array should be processed. The only solution i see is:
function expandArray(objects, callback){
number_of_requests=objects.length-1;
for(i in objects){
$.getJSON(request,function(){
//expanding array
if(--number_of_reuests==0){
callback();
}
});
}
}
But as requests are executed in parallel there is chance of race condition. Variable number_of_requests can be edited by two "threads" simultaneously. How to avoid chance of race condition?
Is it possible to rework your AJAX so that it all goes in one request? That part of the system will be the biggest bottleneck and can get tricky (as you've found out), so the less requests you make, the better.
The only other way I could think would be if each request mutated its related object, setting a flag value or something, and then you looped through all the objects to check if all the flags had been set yet.
Isn't Javascript single threaded? The kind of condition you talk of wouldn't occur.
It's more complex, but I would suggest using a third function to monitor the results. Something like this:
1) Start monitor - using an appropriate interval monitor (using an interval to test here is important - a simple loop would lock the JS engine up tight) the results of requests A and B.
2) Call request A.
3) Call request B.
4) Request B finishes - callback function sets a "B All Done!" value.
5) Request A finishes - callback function sets an "A All Done!" value.
6) Monitor recognizes that both requests are completed and calls function which uses data from both.
Worded more simply this is "dependencies" (multiple calls)to "monitor" (the function that checks dependencies) to "completion" (the action to take when all is ready).
You can create the callbacks and the completion function as nested functions to maintain encapsulation and reduce "little" in the global scope.
You can extend your dependencies as far as you like.