关于项目自动化测试架构的改良计划 - 解析XInclude标记

因为在test_suite.xml中，我们多处使用了XInclude标记，他们会被申明在一个叫"http://www.w3.org/2001/XInclude" 的名字空间中，并且引入部分用xi:include来声明，我们这个类的作用就是把这些所有的<xi:include>的部分，都用被引入的文件插入和替换掉。

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

/**   * This class will handle converting a xinclude+xpointer marked xml file to a normal xml file   * Because of the shortage of the current jdk ,it only support the xPointer with element instead of NCName   *@author cwang58   *@created date: Jun 10, 2013   */ public  class  XIncludeConverter {                                                                                                                                                          /**       * this method will handle change the XInclude+XPointer marked xml as normal xml       * @param origialXML the original xml which has the xInclude feature       * @return the parsedXML without the xInclude feature       */      public  static  String convertXInclude2NormalXML(String originalXML)              throws  SAXException,ParserConfigurationException,TransformerConfigurationException,IOException,TransformerException{            DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();            //open up the namespace aware so that it can recognize the "xi" namespace            factory.setNamespaceAware( true );            //let this SAXParser support XInclude            factory.setXIncludeAware( true );            //factory.setValidating(true);            //ignore all the comments added in the source document            factory.setIgnoringComments( true );            DocumentBuilder docBuilder = factory.newDocumentBuilder();            Document doc = docBuilder.parse( new  InputSource( new  ByteArrayInputStream(originalXML.getBytes( "utf-8" ))));            Transformer transformer = TransformerFactory.newInstance().newTransformer();            //format the output xml string so it support indent and more readable            transformer.setOutputProperty(OutputKeys.INDENT,  "yes" );            //initialize StreamResult with File object to save to file            StreamResult result =  new  StreamResult( new  StringWriter());            DOMSource source =  new  DOMSource(doc);            transformer.transform(source, result);            return  result.getWriter().toString();          }

这里讲一个小插曲，其实，W3C中，XInclude经常和Xpointer联合起来应用的，xpointer可以帮助来定位目标文件的某个小片段而不是整个目标文件，定位方法可以用element(),或者xpointer()，如果element的话，可以用（/1/2/3)这种方式来定位DOM，或者基于 id，对应java的解析框架是xerce,但是非常不幸运的是，最新版本的xerce框架只支持element(/1/3/4/5)这种定位，而对于基于schema-id的方式，也就是某个element声明了id的情况，它没办法定位，但是未来可能会支持这个功能。

http://xerces.apache.org/xerces2-j/faq-xinclude.html#faq-8

基于上述的局限性，我决定只采用xi:include来包含全部文件，然后局部调整的做法，并且绕过xpointer。

所以实现代码如上所示，事实上从JDK 1.6开始，他已经提供了对XInclude的支持，内部是委托给xerce来实现的，这是对应架构图的第3-4步骤。

本文转自 charles_wang888 51CTO博客，原文链接：http://blog.51cto.com/supercharles888/1221731，如需转载请自行联系原作者