xercesJ for XSD 1.1：验证器正确处理断言，SAXParser显然没有

我正在为我的java项目构建一个带有xerces-2_11_0-xml-schema-1.1-beta库的xml验证器，以支持XSD 1.1特定的功能，如断言 .

考虑这个架构

<?xml version="1.1" encoding="UTF-8"?>
<schema targetNamespace="http://www.example.org/Example" elementFormDefault="qualified" xmlns="http://www.w3.org/2001/XMLSchema" xmlns:tns="http://www.example.org/Example">

    <complexType name="NodeType">
        <sequence>
            <element name="Node" type="tns:NodeType"
                maxOccurs="unbounded" minOccurs="0">
            </element>
        </sequence>
        <attribute name="partnumber">
            <simpleType>
                <restriction base="string">
                    <pattern value="[A-Z0-9_\-]+"></pattern>
                </restriction>
            </simpleType>
        </attribute>
        <assert test="starts-with(@partnumber,../@partnumber)"/>
    </complexType>

    <element name="Node" type="tns:NodeType"></element>
</schema>

我开始使用“javax.xml.validation.Validator”类

...
StreamSource xmlSource = new StreamSource(new File("example.xml"));

SchemaFactory sf = SchemaFactory.newInstance("http://www.w3.org/XML/XMLSchema/v1.1");
Schema s = sf.newSchema(new File("example.xsd"));

Validator v = s.newValidator();
v.setErrorHandler(new MyErrorHandler())
v.validate(xmlSource)
...

private static class MyErrorHandler extends DefaultHandler {

...
        public void error(SAXParseException e) throws SAXException {
            System.out.println("Error: "); 
            System.out.println("   Public ID: "+e.getPublicId());
            System.out.println("   System ID: "+e.getSystemId());
            System.out.println("   Line number: "+e.getLineNumber());
            System.out.println("   Column number: "+e.getColumnNumber());
            System.out.println("   Message: "+e.getMessage());;
        }
...
}

此解决方案有效：example.xsd中的断言测试得到正确处理，xml文件的验证运行顺利（正确评估断言测试）

然后我用SAXParser替换Validator（原因：更好地控制sax解析阶段）

...
SchemaFactory sf = SchemaFactory.newInstance("http://www.w3.org/XML/XMLSchema/v1.1");
Schema s = sf.newSchema(new File("example.xsd"));

SAXParserFactory spf = SAXParserFactory.newInstance();
spf.setSchema(s);
spf.setNamespaceAware(true);
spf.setValidating(true);
spf.setFeature("http://apache.org/xml/features/validation/schema", true);

SAXParser parser = spf.newSAXParser();
XMLReader r = parser.getXMLReader();
r.setErrorHandler(new MyErrorHandler())
r.parse("example.xml");
...

在第二种情况下，解析阶段在xsd“assert”子句上遇到错误，并显示以下消息：

s4s-elt-invalid-content.1: The content of 'NodeType' is invalid.  Element 'assert' is invalid, misplaced, or occurs too often.

如果我删除example.xsd中的assert子句（从而使其符合1.0），验证运行正确，所以我认为SAXParser仍在使用1.0规范 .

我是否遗漏了配置SAXParserFactory或SAXParser本身以便坚持1.1架构规则的内容？