Possible bugs

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
1 message Options
Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Possible bugs

Tim Mills-2
I'd be grateful for a response on the possible bugs in the XML
conformance test suite listed below.

Possible bug 1
============

The base URI for one group of tests in xmlconf.xml is incorrect.

The following:

<TESTCASES xml:base="eduni/namespaces/misc/">
     &eduni-misc;
</TESTCASES>

should be:

<TESTCASES xml:base="eduni/misc/">
     &eduni-misc;
</TESTCASES>


Possible Bug 2
============

Test "rmt-e3e-13" is marked as TYPE="invalid".  According to the
documentation

'Non-validating parsers must also accept "invalid" testcases,  but
validating ones must reject them.'

so:

1.  A validating parser will reject E13.xml (below) as invalid.
2.  A non-validating parser will accept E13.xml (below) as well- formed XML.

However the comment for this test case

<TEST RECOMMENDATION="XML1.0-errata3e" SECTIONS="E13" URI="E13.xml"
ID="rmt-e3e-13" TYPE="invalid">
Even internal parameter entity references are enough to make undeclared
entities into mere validity errors rather than well-formedness errors.
</TEST>

indicates that

1.  A validating parser will reject E13.xml as invalid due to the
undeclared entity ent2.
2.  A non-validating parser will reject E13.xml as not well formed due
to the undeclared entity ent2.

If I'm not mistaken, this doesn't match the intent of TYPE="invalid".

E13.xml

<!-- Even internal parameter entity references are enough to make
undeclared entities into mere validity errors rather than
well-formedness errors. -->
<!DOCTYPE foo [
<!ENTITY % pe "<!ENTITY ent1 'text'>">
%pe;
<!ELEMENT foo ANY>
]>
<foo>&ent2;</foo>


Loading...