Bug 70501

Created attachment 111776 [details] Patch

Comment on attachment 111776 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=111776&action=review > LayoutTests/fast/dom/MicroData/006-expected.txt:5 > PASS document.getItems(' http://example.com/foo http://example.com/bar ').length == 2 is true > -PASS document.getItems(' http://example.com/foo data:text/plain ').length == 2 is true > -PASS document.getItems(' http://example.com/foo data:text/plain http://example.com/foo').length == 2 is true > +PASS document.getItems(' http://example.com/foo http://example.com/bar http://example.com/foo').length == 2 is true Why the change? > LayoutTests/fast/dom/MicroData/006.html:11 > -<div itemscope itemtype="http://example.com/foo"></div> > -<div itemscope itemtype="http://example.com/bar"> > -<div itemscope itemtype="data:text/plain"></div> > +<div itemscope itemtype="http://example.com/foo http://example.com/bar"></div> > +<div itemscope itemtype="http://example.com/bar http://example.com/foo"> Ditto. > LayoutTests/fast/dom/MicroData/006.html:16 > -shouldBeTrue("document.getItems(' http://example.com/foo data:text/plain ').length == 2"); > -shouldBeTrue("document.getItems(' http://example.com/foo data:text/plain http://example.com/foo').length == 2"); > +shouldBeTrue("document.getItems(' http://example.com/foo http://example.com/bar http://example.com/foo').length == 2"); Ditto. > LayoutTests/fast/dom/MicroData/itemtype-add-remove-tokens.html:35 > +debug("<br>itemType.remove must not make any changes if a non-existing token is removed."); "non-existing token is removed" sounds odd. How about "remove is called for a token that doesn't exist"?

(In reply to comment #2) > (From update of attachment 111776 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=111776&action=review > > > LayoutTests/fast/dom/MicroData/006.html:11 > > -<div itemscope itemtype="http://example.com/foo"></div> > > -<div itemscope itemtype="http://example.com/bar"> > > -<div itemscope itemtype="data:text/plain"></div> > > +<div itemscope itemtype="http://example.com/foo http://example.com/bar"></div> > > +<div itemscope itemtype="http://example.com/bar http://example.com/foo"> > > Why the change? Accroding to the latest spec getItems() should return all the elements that are each top-level microdata items whose types include "all the types specified" in the method's argument. Earlier the spec specified that getItems() should return all the elements that are each top-level microdata items whose types include "any one of the types specified" in the method's argument. So if we do not modify the below test-case, -<div itemscope itemtype="http://example.com/foo"></div> -<div itemscope itemtype="http://example.com/bar"> +<div itemscope itemtype="http://example.com/foo http://example.com/bar"></div> +<div itemscope itemtype="http://example.com/bar http://example.com/foo"> document.getItems(' http://example.com/foo http://example.com/bar ').length == 2 will always be false. Hence I made this changes. And I have modified this test-case more perfect in my new patch. > > > LayoutTests/fast/dom/MicroData/itemtype-add-remove-tokens.html:35 > > +debug("<br>itemType.remove must not make any changes if a non-existing token is removed."); > > "non-existing token is removed" sounds odd. How about "remove is called for a token that doesn't exist"? Done!

Created attachment 112305 [details] Updated patch

Comment on attachment 112305 [details] Updated patch Clearing flags on attachment: 112305 Committed r98353: <http://trac.webkit.org/changeset/98353>

All reviewed patches have been landed. Closing bug.