[dsdl-discuss] Re: Schematron question: 5.4.6

From: Rick Jelliffe <rjelliffe@allette.com.au>
Date: Sun Jan 21 2007 - 06:28:44 UTC

MURATA Makoto said:
>> I think that " that is the name of a node" should be removed from this
>> sentence and "is" should be replaced with "shall". This sentence is
>> normative. And an informative note stating the intention (i.e., it
>> should return the name of a node) should be added.
>
> The problem of the current wording is that full-conformant
> implementations are required to detect what cannot be easily detected.
>
> In the conformance section, you wrote:
>
> A full-conformance implementation shall be able to determine for
> any XML document whether it is a correct schema.
>
> That is, if a malicious schema specifies a name query that returns "$$$$",
> full-conformance implementions must detect this error.

There are two levels of conformance. S7.1 Simple conformance is when it
follows the RELAX NG grammar. s7.2 Full conformance has many more
constraints, such as checking the grammar of the queries.

However, s7.2 NOTE 1 says that "It is not a requirement ... that a full
conformance implementation shall be able to detect ... whether the queries
are feasible against some schema language for the instance."

This covers the case of a name query that returns a string. A full
conformance implementation is not required to detect this. It would be a
runtime error or the implementation could have special code to handle it.

Cheers
Rick Jelliffe

Cheers
Rick

--
DSDL members discussion list
To unsubscribe, please send a message with the
command  "unsubscribe" to dsdl-discuss-request@dsdl.org
(mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)
Received on Sun Jan 21 07:33:54 2007

This archive was generated by hypermail 2.1.8 : Sun Jan 21 2007 - 10:23:02 UTC