java xpath解析html_java使用xpath和dom4j解析xml

packageXPath;

importjava.io.IOException;

importjava.io.InputStream;

importjava.util.HashMap;

importjava.util.List;

importjava.util.Map;

importjavax.xml.parsers.DocumentBuilder;

importjavax.xml.parsers.DocumentBuilderFactory;

importjavax.xml.parsers.ParserConfigurationException;

importjavax.xml.xpath.XPathConstants;

importjavax.xml.xpath.XPathExpression;

importjavax.xml.xpath.XPathExpressionException;

importjavax.xml.xpath.XPathFactory;

importorg.dom4j.Document;

importorg.dom4j.DocumentException;

importorg.dom4j.Element;

importorg.dom4j.XPath;

importorg.dom4j.io.SAXReader;

importorg.w3c.dom.NodeList;

importorg.xml.sax.SAXException;

/**

*DOM4JDOMXMLXPath

*/

publicclassTestDom4jXpath{

publicstaticvoidmain(String[]args){

read1();

read2();

read3();

read4();//read3()方法一样，但是XPath表达式不同

read5();

}

publicstaticvoidread1(){

/*

*uselocal-name()andnamespace-uri()inXPath

*/

try{

longstartTime=System.currentTimeMillis();

SAXReaderreader=newSAXReader();

InputStreamin=TestDom4jXpath.class.getClassLoader().getResourceAsStream("XPath\\XXX.xml");

Documentdoc=reader.read(in);

/*Stringxpath="//*[local-name()='Workbook'andnamespace-uri()='urn:schemas-microsoft-com:office:spreadsheet']"

+"/*[local-name()='Worksheet']"

+"/*[local-name()='Table']"

+"/*[local-name()='Row'][4]"

+"/*[local-name()='Cell'][3]"

+"/*[local-name()='Data'][1]";*/

Stringxpath="//*[local-name()='Row'][4]/*[local-name()='Cell'][3]/*[local-name()='Data'][1]";

System.err.println("=====uselocal-name()andnamespace-uri()inXPath====");

System.err.println("XPath："+xpath);

@SuppressWarnings("unchecked")

Listlist=doc.selectNodes(xpath);

for(Objecto:list){

Elemente=(Element)o;

Stringshow=e.getStringValue();

System.out.println("show="+show);

longendTime=System.currentTimeMillis();

System.out.println("程序运行时间："+(endTime-startTime)+"ms");

}

}catch(DocumentExceptione){

e.printStackTrace();

}

}

publicstaticvoidread2(){

/*

*setxpathnamespace(setNamespaceURIs)

*/

try{

longstartTime=System.currentTimeMillis();

Mapmap=newHashMap();

map.put("Workbook","urn:schemas-microsoft-com:office:spreadsheet");

SAXReaderreader=newSAXReader();

InputStreamin=TestDom4jXpath.class.getClassLoader().getResourceAsStream("XPath\\XXX.xml");

Documentdoc=reader.read(in);

Stringxpath="//Workbook:Row[4]/Workbook:Cell[3]/Workbook:Data[1]";

System.err.println("=====usesetNamespaceURIs()tosetxpathnamespace====");

System.err.println("XPath："+xpath);

XPathx=doc.createXPath(xpath);

x.setNamespaceURIs(map);

@SuppressWarnings("unchecked")

Listlist=x.selectNodes(doc);

for(Objecto:list){

Elemente=(Element)o;

Stringshow=e.getStringValue();

System.out.println("show="+show);

longendTime=System.currentTimeMillis();

System.out.println("程序运行时间："+(endTime-startTime)+"ms");

}

}catch(DocumentExceptione){

e.printStackTrace();

}

}

publicstaticvoidread3(){

/*

*setDocumentFactory()namespace(setXPathNamespaceURIs)

*/

try{

longstartTime=System.currentTimeMillis();

Mapmap=newHashMap();

map.put("Workbook","urn:schemas-microsoft-com:office:spreadsheet");

SAXReaderreader=newSAXReader();

InputStreamin=TestDom4jXpath.class.getClassLoader().getResourceAsStream("XPath\\XXX.xml");

reader.getDocumentFactory().setXPathNamespaceURIs(map);

Documentdoc=reader.read(in);

Stringxpath="//Workbook:Row[4]/Workbook:Cell[3]/Workbook:Data[1]";

System.err.println("=====usesetXPathNamespaceURIs()tosetDocumentFactory()namespace====");

System.err.println("XPath："+xpath);

@SuppressWarnings("unchecked")

Listlist=doc.selectNodes(xpath);

for(Objecto:list){

Elemente=(Element)o;

Stringshow=e.getStringValue();

System.out.println("show="+show);

longendTime=System.currentTimeMillis();

System.out.println("程序运行时间："+(endTime-startTime)+"ms");

}

}catch(DocumentExceptione){

e.printStackTrace();

}

}

publicstaticvoidread4(){

/*

*同read3()方法一样，但是XPath表达式不同

*/

try{

longstartTime=System.currentTimeMillis();

Mapmap=newHashMap();

map.put("Workbook","urn:schemas-microsoft-com:office:spreadsheet");

SAXReaderreader=newSAXReader();

InputStreamin=TestDom4jXpath.class.getClassLoader().getResourceAsStream("XPath\\XXX.xml");

reader.getDocumentFactory().setXPathNamespaceURIs(map);

Documentdoc=reader.read(in);

Stringxpath="//Workbook:Worksheet/Workbook:Table/Workbook:Row[4]/Workbook:Cell[3]/Workbook:Data[1]";

System.err.println("=====usesetXPathNamespaceURIs()tosetDocumentFactory()namespace====");

System.err.println("XPath："+xpath);

@SuppressWarnings("unchecked")

Listlist=doc.selectNodes(xpath);

for(Objecto:list){

Elemente=(Element)o;

Stringshow=e.getStringValue();

System.out.println("show="+show);

longendTime=System.currentTimeMillis();

System.out.println("程序运行时间："+(endTime-startTime)+"ms");

}

}catch(DocumentExceptione){

e.printStackTrace();

}

}

publicstaticvoidread5(){

/*

*DOMandXPath

*/

try{

longstartTime=System.currentTimeMillis();

DocumentBuilderFactorydbf=DocumentBuilderFactory.newInstance();

dbf.setNamespaceAware(false);

DocumentBuilderbuilder=dbf.newDocumentBuilder();

InputStreamin=TestDom4jXpath.class.getClassLoader().getResourceAsStream("XPath\\XXX.xml");

org.w3c.dom.Documentdoc=builder.parse(in);

XPathFactoryfactory=XPathFactory.newInstance();

javax.xml.xpath.XPathx=factory.newXPath();

//选取所有class元素的name属性

Stringxpath="//Workbook/Worksheet/Table/Row[4]/Cell[3]/Data[1]";

System.err.println("=====DomXPath====");

System.err.println("XPath："+xpath);

XPathExpressionexpr=x.compile(xpath);

NodeListnodes=(NodeList)expr.evaluate(doc,XPathConstants.NODE);

for(inti=0;i

System.out.println("show="+nodes.item(i).getNodeValue());

longendTime=System.currentTimeMillis();

System.out.println("程序运行时间："+(endTime-startTime)+"ms");

}

}catch(XPathExpressionExceptione){

e.printStackTrace();

}catch(ParserConfigurationExceptione){

e.printStackTrace();

}catch(SAXExceptione){

e.printStackTrace();

}catch(IOExceptione){

e.printStackTrace();

}

}

}