On Jul 22, 2008, at 10:26 AM, Benjamin Hawkes-Lewis wrote: > Dan Wood wrote: >> Yeah, that's one approach but it is way overkill for many needs, and >> invokes threads that aren't what you want for a simple function, and >> also only go from named entity into rendered Unicode character, and >> not the opposite (as you mention). I want the named entities rather >> than numbered for human-coder-readability.... > > Note that for the vast majority of characters in Unicode, no named > entities exist.
I imagine Dan is aware that the named references are a small subset of all Unicode characters. However, I can understand why that tiny subset of Unicode characters might prove useful as character references, making the generated source more human readable (especially for those characters that are part of the Unicode 'common' script). Having said that, It seems to me this sounds like a job for HTMLTidy. It might make sense to approach the HTMLTidy group to implement this. If HTMLTidy already performed this conversion (as an option), WebKit could easily add a method to expose that functionality when serializing the DOMDocument. Considering how ubiquitous HTMLTidy is, that makes it an even better place for code reuse. Take care, Rob _______________________________________________ webkit-dev mailing list [email protected] http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
