Re: XHTML character entity support

On Tue, Nov 3, 2009 at 10:34 AM, Henri Sivonen <hsivonen@iki.fi> wrote:
> On Nov 3, 2009, at 17:06, Shelley Powers wrote:
>
>> I'm going to focus on the relevant part of this discussion to the HTML
>> WG: there are rules defined for how undefined entities are handled,
>> and these rules defined in the XML specification. There may be some
>> issues of interpretation, but such issues are specific to the XML
>> spec, not the HTML5 spec.
>>
>> As such, no further explanations or additional specifications are
>> necessary in HTML5.
>>
>> Am I correct in this?
>
> Not in my opinion. If predictably uniform behavior between UAs is wanted and
> if we want to make it non-mysterious for implementors how to performantly
> parse application/xhtml+xml content written for browsers, this WG should
> specify normative entity resolver behavior (i.e. mappings from public id and
> system id pairs onto streams).

First, application/xhtml+xml is not written for browsers. Browsers may
be the biggest implementors, but they're not the only implementor.

Second, what you're discussing seems to be something that's general to
all XML parsers, not just browsers. As such, it would be better
defined as part of XML core, rather than an end case like HTML5. Or is
there something unique that the browsers do that falls outside of
normal XML parsing?

Besides, the point is moot: XHTML5 does not have a DTD, only the five
predefined works with the XHTML.

The job of this working group is not to normalize all of the browser
quirks and differences. Don't you agree?


>
> As a practical matter, if I'm using SAX in Java, I can't get a browser-style
> EntityResolver off-the-shelf as part of a common org.apache package. (Or
> maybe I could but I'm unaware.)
>

What exactly would you expect the EntityResolver to do with XHTML that
it wouldn't do with other flavors of XML?

Shelley


> --
> Henri Sivonen
> hsivonen@iki.fi
> http://hsivonen.iki.fi/
>
>
>

Received on Tuesday, 3 November 2009 20:07:14 UTC