I would much prefer this API change to resorting to caching. Caching worked great in the days of static web pages. Today we are just asking for trouble.
cheers, David On 18/04/10 11:58 PM, Alexander Surkov wrote: > Hi. > > IAccessibleHypertext is not performant way in Firefox to get all > hyperlinks and their offsets for the hypertext accessible. Hyperlinks > and their offset could be cached within hypertext accessible but it's > might be a pain to keep the cache updated and it might be not quick > since we need to update the cache whenever the characters of child > text node or child nodes are inserted/deleted. Could we consider a > method allowing to get all hyperlinks and their hypertext offsets all > together? For example, > > HRESULT hyperLinks ([out, size_is( ,nLinkCount, )] IUnknown ***links, > [out, size_is( ,nLinkCount, )] > long **linkIndexes, > [out, retval] long *nLinkCount) > > This method should be much quicker when web site contains huge amount > of links and it should be friendly to out-of-process screen readers. > > Also we could return an array of pairs (link, linkIndex) if this looks > more nice than two arrays. > > Thank you. > Alex. > _______________________________________________ > Accessibility-ia2 mailing list > [email protected] > https://lists.linux-foundation.org/mailman/listinfo/accessibility-ia2 > _______________________________________________ Accessibility-ia2 mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/accessibility-ia2
