W3C home > Mailing lists > Public > public-xml-core-wg@w3.org > April 2016

Re: Proposed changes to XInlude test suite

From: Paul Grosso <paul@paulgrosso.name>
Date: Fri, 15 Apr 2016 14:38:12 -0500
To: public-xml-core-wg@w3.org
Message-ID: <57114324.4000301@paulgrosso.name>


On 2016-04-15 10:36, Norman Walsh wrote:
> Hello,
>
> I’m working my way through the test suite, fixing bugs in my implementation (in
> XML Calabash). I assert that the proposed results for
>
> https://github.com/w3c/xinclude/blob/testsuite11/tests/Imaq/result/XInclude/pex6a.xml
>
> and
>
> https://github.com/w3c/xinclude/blob/testsuite11/tests/Imaq/result/XInclude/pex6b.xml
>
> are incorrect. In each case, there is a newline in the XIncluded document and a
> newline following the xi:include element itself. I propose that the correct result
> has two newlines before the </x> element.
>
> I also propose to remove this test:
>
> https://github.com/w3c/xinclude/blob/testsuite11/tests/Nist/test/docs/nist-include-03.xml
>
> The supported values of the parse attribute are no longer limited to ‘xml’ and ‘text’.
> The set is implementation defined and an unrecognized value is a recoverable error
> in XInclude 1.1.
>

Well, I think it's still a valid test.  Assuming no implementation
recognizes the value "xtext", and given that the xinclude element
has no fallback, the result should (still) be a fatal error, right?

(I do note that we have a "bad-parse" test that checks that an
unrecognized value is a recoverable error, so that's good.)

paul
Received on Friday, 15 April 2016 19:43:57 UTC

This archive was generated by hypermail 2.3.1 : Friday, 15 April 2016 19:43:58 UTC