Summary: | [Streams API] Implement IsReadableStreamDisturbed according to spec | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Xabier Rodríguez Calvar <calvaris> | ||||
Component: | New Bugs | Assignee: | Xabier Rodríguez Calvar <calvaris> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | benjamin, commit-queue, darin, youennf | ||||
Priority: | P2 | ||||||
Version: | WebKit Nightly Build | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Xabier Rodríguez Calvar
2015-11-17 12:33:21 PST
Created attachment 265697 [details]
Patch
Comment on attachment 265697 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=265697&action=review > Source/WebCore/testing/Internals.idl:413 > + [Conditional=STREAMS_API, CallWith=ScriptState] boolean isReadableStreamDisturbed(any stream); Is there really no indirect way to test this “disturbed” state? That’s what we should be doing. I think exposing the internal disturbed flag is an inferior way to do the testing. The specification does not actually require a “disturbed” flag; just that everything works “as if” there is one. I suppose we are inheriting this testing strategy from whoever is creating the test suite, though. (In reply to comment #2) > Is there really no indirect way to test this “disturbed” state? That’s what > we should be doing. I think exposing the internal disturbed flag is an > inferior way to do the testing. The specification does not actually require > a “disturbed” flag; just that everything works “as if” there is one. I > suppose we are inheriting this testing strategy from whoever is creating the > test suite, though. Yes, I am doing it like this precisely because there are such tests. Actually it is kinda surprinsing that they exist in such way. Anyway, the flag is exposed thru the test internals so never to the user in the browser. AFAIK, disturbed flag is only needed by fetch API atm. It will be testable as usual once fetch API becomes available to WebKit which I hope will happen in a near future. The reference testsuite is built to ensure spec and ref impl work well. Comment on attachment 265697 [details] Patch Clearing flags on attachment: 265697 Committed r192621: <http://trac.webkit.org/changeset/192621> All reviewed patches have been landed. Closing bug. |