Our structured cloning implementation does not encode all of RegExp's flags

• Component: Bindings
• Status: RESOLVED FIXED

[Chris Dumez]

Our structured cloning implementation does not encode all of RegExp's flags.

This is causing WebKit to fail one of the subtests on http://wpt.live/IndexedDB/structured-clone.any.html?61-80. This subtest is passing in both Chrome and Firefox.

[Chris Dumez]

Created attachment 433093 [details]
Patch

[Yusuke Suzuki]

Comment on attachment 433093 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=433093&action=review

> Source/WebCore/bindings/js/SerializedScriptValue.cpp:1256
>                      flags[flagCount++] = 'i';
>                  if (regExp->regExp()->multiline())
>                      flags[flagCount++] = 'm';
> +                if (regExp->regExp()->dotAll())
> +                    flags[flagCount++] = 's';
> +                if (regExp->regExp()->unicode())
> +                    flags[flagCount++] = 'u';
> +                if (regExp->regExp()->sticky())
> +                    flags[flagCount++] = 'y';

Need to add `hasIndices` flag ("d") too (this is new flag introduced super recently).
Also, can you add a test for "d" flag case?

[Chris Dumez]

(In reply to Yusuke Suzuki from comment #2)
> Comment on attachment 433093 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=433093&action=review
> 
> > Source/WebCore/bindings/js/SerializedScriptValue.cpp:1256
> >                      flags[flagCount++] = 'i';
> >                  if (regExp->regExp()->multiline())
> >                      flags[flagCount++] = 'm';
> > +                if (regExp->regExp()->dotAll())
> > +                    flags[flagCount++] = 's';
> > +                if (regExp->regExp()->unicode())
> > +                    flags[flagCount++] = 'u';
> > +                if (regExp->regExp()->sticky())
> > +                    flags[flagCount++] = 'y';
> 
> Need to add `hasIndices` flag ("d") too (this is new flag introduced super
> recently).
> Also, can you add a test for "d" flag case?

Will do, thanks for pointing that out.

[Chris Dumez]

Created attachment 433105 [details]
Patch

[Yusuke Suzuki]

Comment on attachment 433105 [details]
Patch

r=me

[EWS]

Committed r279706 (239498@main): <https://commits.webkit.org/239498@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 433105 [details].

[Radar WebKit Bug Importer]

<rdar://problem/80301211>