Bug 223071 – Implement AbortSignal.abort()

• Component: DOM
• Status: RESOLVED FIXED

[James M Snell]

https://github.com/whatwg/dom/pull/960 introduces a new utility `AbortSignal.abort()` that creates and returns an already aborted AbortSignal.

Example implementation in Node.js: https://github.com/nodejs/node/pull/37693. 

WPT Test: https://github.com/web-platform-tests/wpt/pull/28003

[Radar WebKit Bug Importer]

<rdar://problem/75575483>

[Chris Dumez]

Created attachment 423772 [details]
Patch

[EWS Watchlist]

This patch modifies the imported WPT tests. Please ensure that any changes on the tests (not coming from a WPT import) are exported to WPT. Please see https://trac.webkit.org/wiki/WPTExportProcess

[Darin Adler]

Comment on attachment 423772 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=423772&action=review

> Source/WebCore/dom/AbortSignal.h:45
> +    static Ref<AbortSignal> createAborted(ScriptExecutionContext&);

Why not name this abort to match the DOM public API?

[Chris Dumez]

Comment on attachment 423772 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=423772&action=review

>> Source/WebCore/dom/AbortSignal.h:45
>> +    static Ref<AbortSignal> createAborted(ScriptExecutionContext&);
> 
> Why not name this abort to match the DOM public API?

Because the abort() name is very confusing for such operation. Also, AbortSignal already has an abort() member function to actually abort the signal. I wish the spec had called this aborted() instead of abort()..

[Sam Weinig]

Comment on attachment 423772 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=423772&action=review

>>> Source/WebCore/dom/AbortSignal.h:45
>>> +    static Ref<AbortSignal> createAborted(ScriptExecutionContext&);
>> 
>> Why not name this abort to match the DOM public API?
> 
> Because the abort() name is very confusing for such operation. Also, AbortSignal already has an abort() member function to actually abort the signal. I wish the spec had called this aborted() instead of abort()..

I think we should match the spec here and rename the other function to signalAbort, which is what the spec calls it. Even if we don’t love the names, it’s much easier if we match things when we can.

[EWS]

commit-queue failed to commit attachment 423772 [details] to WebKit repository. To retry, please set cq+ flag again.

[Chris Dumez]

(In reply to Sam Weinig from comment #6)
> Comment on attachment 423772 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=423772&action=review
> 
> >>> Source/WebCore/dom/AbortSignal.h:45
> >>> +    static Ref<AbortSignal> createAborted(ScriptExecutionContext&);
> >> 
> >> Why not name this abort to match the DOM public API?
> > 
> > Because the abort() name is very confusing for such operation. Also, AbortSignal already has an abort() member function to actually abort the signal. I wish the spec had called this aborted() instead of abort()..
> 
> I think we should match the spec here and rename the other function to
> signalAbort, which is what the spec calls it. Even if we don’t love the
> names, it’s much easier if we match things when we can.

Ok, will update the name.

[Chris Dumez]

Created attachment 423885 [details]
Patch

[EWS]

Committed r274773: <https://commits.webkit.org/r274773>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 423885 [details].