[LC-2197] RE: Complex Datatype representation map

Hi Suzuki-san,

Thank you for providing a feedback that allows the WG to improve
the language of the spec in a way to make it clearer to all the readers.

As you alluded in your comment, the types that are elligible are
only simple types. Therefore, the implementors of user-defined
datatype representations not only do not need to be concerned about
event-codes chore, but also indeed have no access to them.

We acknowledge that some of the sentences in that section can be
improved for clarity, as well as Example 7-3 in which the type
name "geo:geometricSurface" might have indicated that it represented
a complex type thus was misleading. We will make changes to make it
unequivocal both in the language and the example.

Thanks!

-taki


-----Original Message-----
From: public-exi-request@w3.org [mailto:public-exi-request@w3.org] On Behalf Of Gengo Suzuki
Sent: Monday, April 20, 2009 5:42 PM
To: public-exi@w3.org
Subject: Complex Datatype representation map

Hello,

I have a question about Datatype Representation Map

In 7.4 of Working Draft, Example 7-2 shows single type representation
map. Below the example an outline to implement it are described, and
I can image to use it.

Example 7-3 shows (perhaps) representation map for complex type.
There is no detailed explanation about it.

I feel that it is not trivial to implement complex type map, because
  in a complex type case event code determination rules are unclear
to treat child types. Mapping influence grammar production.

Is it responsibility for implementor of EXI processor?
Or can datatype representation map be used only in single type
case?

Sincerely,

//---------------------------------------------------------------
NTT Cyber Space Laboratories
Gengo Suzuki <suzuki.gengo@lab.ntt.co.jp>
TEL: +81-46-859-3412 FAX: +81-46-859-2768
----------------------------------------------------------------//

Received on Wednesday, 29 April 2009 17:57:36 UTC