Re: attribute order (RE: Syntax Sugar and XML information models)

["Martin v. Loewis" <martin@loewis.home.cs.tu-berlin.de>]

> Is there some kind of processing advantage for _parsers_ (not
> applications) to report attribute values in any sequence other than
> their original sequence?

Parsers can use dictionaries to report the attributes, which are more
efficient to access when given a key. Not sure whether this is a
processing advantage on the parser or on the application side, but it
should allow for faster overall processing.

Now, parsers could preserve the original order side-by-side with the
dictionary. In that case, not preserving the order simplifies the
implementation of the parser. Ordered dictionaries with efficient
lookup and insertion are rarely found as predefined types.

Regards,
Martin