As part of the larger <style scoped> implementation, implement the necessary changes to the DOM and HTML classes.
Created attachment 106763 [details] Patch The patch adds the 'scoped' attribute to the HTMLStyleElement IDL and HTMLAttributeNames.in, and tests setting and re-setting the attribute in various circumstances.
Committed r94784: <http://trac.webkit.org/changeset/94784>
Rolled out in r106777 in order to allow for more discussion
Comment on attachment 106763 [details] Patch Clearing flags on attachment: 106763 Committed r94809: <http://trac.webkit.org/changeset/94809>
All reviewed patches have been landed. Closing bug.
Reverted r94809 for reason: commit _still_ premature, despite CQ eagerness Committed r94813: <http://trac.webkit.org/changeset/94813>
Do we need an ENABLE flag for this, at least while the implementation is incomplete?
Opened a new bug https://bugs.webkit.org/show_bug.cgi?id=72848 for the ENABLE flag.
A question is, do we need to add the flag to code that handles the attribute/property, but has no outside side effects? I.e: - attribute reflection code - registration with parent element - window.internals.numberOfScopedHTMLStyleChildren function
(In reply to comment #9) > A question is, do we need to add the flag to code that handles the attribute/property, but has no outside side effects? I.e: > > - attribute reflection code > - registration with parent element > - window.internals.numberOfScopedHTMLStyleChildren function Yes. One of the uses of the flag is to remove a feature in case it is unstable, and that wouldn't work if we only limited the flag to the author-visible effects.
(In reply to comment #10) > Yes. One of the uses of the flag is to remove a feature in case it is unstable, and that wouldn't work if we only limited the flag to the author-visible effects. In that case, how do I best handle the addition/changes to HTMLAttributeNames.in and HTMLStyleElement.idl?
(In reply to comment #11) > (In reply to comment #10) > > Yes. One of the uses of the flag is to remove a feature in case it is unstable, and that wouldn't work if we only limited the flag to the author-visible effects. > > In that case, how do I best handle the addition/changes to HTMLAttributeNames.in and HTMLStyleElement.idl? 1) Add item to HTMLAttributeNames.in 2) Add conditional attribute to HTMLStyleElement.idl In other words, follow existing conventions.
Created attachment 117846 [details] patch with flag Same patch as before, with added ENABLE(STYLE_SCOPED) flag
Comment on attachment 117846 [details] patch with flag Clearing flags on attachment: 117846 Committed r101987: <http://trac.webkit.org/changeset/101987>