xerces模式文档的路径

[liaomingxue]

使用xerces java的最新版本发现xml文档的schema无法指定相对路径！xml文档、schema和程序如下。

如果将schema放到当前目录之下，则顺利通过；
如果将schema与xml文档放在相同目录之下，即指望xerces能在xml文档的目录中找到schema，竟失败！

如果将程序中的parser.parse(inputSource);直接改成parser.parse("E:/projects/java/rms/templates/o.xml");
则是可以的，因为这里的参数将被按照URI进行处理，然后就能找到相对目录下的schema，但这样的话，由于URI处理中文有问题，文件名就不能采用中文了。


xml文档如下：
<?xml version="1.0" encoding="gb2312"?>
<ResourceReg xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNameSpaceSchemaLocation="registration.xsd">
<ResourceFig>
<ResourceKID>512</ResourceKID>
<PortAddr>192.192.192.222:1:1</PortAddr>
<ResourceSID>3</ResourceSID>
</ResourceFig>
<ResourceStatus>
<ZBWZ>135,26</ZBWZ>
<YXZT>1</YXZT>
<CSQB>true</CSQB>
<BKF>2</BKF>
</ResourceStatus>
</ResourceReg>

schema如下：
<?xml version="1.0"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">
<xs:element name="ResourceReg">
<xs:complexType>
<xs:all>
<xs:element name="ResourceFig" type="ResourceFig" minOccurs="1" maxOccurs="1" />
<xs:element name="ResourceStatus" minOccurs="1" maxOccurs="1" />
</xs:all>
</xs:complexType>
</xs:element>
<xs:complexType name="ResourceFig">
<xs:all>
<xs:element name="ResourceKID" type="xs:unsignedShort" />
<xs:element name="PortAddr" type="xs:token" />
<xs:element name="ResourceSID" type="xs:unsignedByte" />
</xs:all>
</xs:complexType>
</xs:schema>

程序如下：
DOMParser parser = new DOMParser();
Document doc = null;
try
{
parser.setFeature("http://xml.org/sax/features/validation",true);
parser.setFeature("http://apache.org/xml/features/validation/schema",true);
InputSource inputSource=new InputSource(
new FileInputStream(
new File("E:/projects/java/rms/templates/o.xml")));
parser.parse(inputSource);
doc = parser.getDocument();
}
catch(Exception e)
{
e.printStackTrace();
}

[AaronLIBo]

用System.getProperty("user.dir")+工用目录

[liaomingxue]

I have found a solution to this problem. I think xcerces should support GB2312 encoding (Chinese).
The solution also solves another problem that xceces does not support files encoded in GB2312.
The solution is as below:

in org.apache.xerces.impl.XMLEntityManager.createReader(InputStream,String,Boolean) add:
/**
* why not supporting GB2312?
* @author [e-mail]
*/
if(encoding.equals("GB2312"))
{
return new InputStreamReader(inputStream,encoding);
}

and in org.apache.xerces.util.URI.initializePath(String, int) update:
else if (!isPathCharacter(testChar))
{
/**
* @author [e-mail]
* The path part of a URI may contain characters which are not included in URI Spec.
*/
if(Character.isUnicodeIdentifierStart(testChar)||Character.isUnicodeIdentifierPart(testChar))
{
++index;
continue;
}

if (testChar == '?' || testChar == '#')
{
break;
}
throw new MalformedURIException("Path contains invalid character: " + testChar);
}

and in org.apache.xerces.util.URI.initializePath(String, int) update:
else if (!isURICharacter(testChar))
{
/**
* A path may contain Chinese characters,
* but I am not sure that the method used here is right.
* And I believe that there must be other parts of this file to be corrected.
* And why not use java.net.URI?
* By [e-mail]
*/
if(Character.isUnicodeIdentifierPart(testChar)||Character.isUnicodeIdentifierStart(testChar))
{
index++;
continue;
}
throw new MalformedURIException(
"Opaque part contains invalid character: " + testChar);
}

[xjjyLOVE]

很好，顶！

[xjjyLOVE]

很好，帮顶！