一键解析XML文件(利用Digester实现可配置)

原创

randy_shandong 2013-04-16 15:36:31 博主文章分类：JAVA ©著作权

文章标签 XML Digester 一键解析 文章分类 Exchange 服务器

©著作权归作者所有：来自51CTO博客作者randy_shandong的原创作品，请联系作者获取转载授权，否则将追究法律责任

大部分程序员，平时工作中除了与Bug相伴之外，想必也会很多种多样的文件打交道吧。当然,XML 就是其中之一，获取交互数据,创建规则等等，都离不开他。XML是个非常强大的描述语言，相比而言,txt之流则功力较弱了些。

XML那么重要，单解析XML的工作却繁杂无聊。原因如下：解析XML工具繁杂,每个人掌握的东西不一样;学习新工具又浪费时间;解析规则只是个体力活。

程序员的辛苦，也就体现于此了。一般程序员总是抱怨忙啊忙，累啊累。我们作为程序员来说，更应该想一想想法设法吧东西提炼一下，减少重复的工作。说这些题外话，是有原因的。我现在的大部分同事，加班加点在哪加班，貌似很辛苦，看一下他们的工作，让人很无奈。代码是一大段一大段的写，确实一晚上写个百十行，是非常“有成就感”的。

如生成XML 文件，用拼串的方式实现。一个小小的xml，往往都50多行。让人很无语。

解析XML, 往往都需要有一套解析规则。而我们可以动态创建解析规则。让后根据规则将xml文件数据放到Map中。思路简单至极。其他扩展，暂时不考虑。

主要步骤分为2部分

第一部分使用Digester实现可配置的规则树结构

1) 规则节点对象

public    class Node{
  private String nodeName;
  private List children = new ArrayList();
  /**
    * 属性列表,以逗号分隔
    */
  private String attrs="";

  public String getAttrs() {
    return attrs;
  }
  public void setAttrs(String values) {
    this.attrs= values;
  }
  public String getNodeName() {
    return nodeName;
  }
  public void setNodeName(String nodeName) {
    this.nodeName = nodeName;
  }
  public List getChildren() {
    return children;
  }
  public void addChildren(Node node) {
    this.children.add(node);
  }
  @Override
  public String toString() {

    return this.nodeName+":"+"have"+this.getChildren().size()+"子节点";
  }
}

(2) 定义xml 解析规则文件，当然也可以写编码实现。

<?xml version="1.0"?>
<!DOCTYPE digester-rules PUBLIC "-//Jakarta Apache //DTD digester-rules XML V1.0//EN" "http://jakarta.apache.org/commons/digester/dtds/digester-rules.dtd">
<digester-rules>
 <pattern value="root">
 <object-create-rule classname="com.aisino.common.parse.Node"/>
 <set-properties-rule />
 <set-next-rule methodname="add"/>

 <pattern value="level1">
 <object-create-rule classname="com.aisino.common.parse.Node"/>
 <set-properties-rule />
 <set-next-rule methodname="addChildren"/>
 <pattern value="level2">
 <object-create-rule classname="com.aisino.common.parse.Node"/>
 <set-properties-rule />
 
 <pattern value="level3">
 <object-create-rule classname="com.aisino.common.parse.Node"/>
 <set-properties-rule />
 <set-next-rule methodname="addChildren"/>
 <pattern value="level4">
 <object-create-rule classname="com.aisino.common.parse.Node"/>
 <set-properties-rule />
 <set-next-rule methodname="addChildren"/>
 <pattern value="level5">
 <object-create-rule classname="com.aisino.common.parse.Node"/>
 <set-properties-rule />
 <set-next-rule methodname="addChildren"/>
 </pattern>
 </pattern>
 </pattern>

 <set-next-rule methodname="addChildren"/>
 </pattern>
 </pattern>
 </pattern>
</digester-rules>

3) 规则解析类RuleLoader

public class RuleLoader {
 private static final Log logger = LogFactory.getLog(RuleLoader.class);
 private URL digesterRulesURL;
 private URL fileURL;
 public RuleLoader(String rulePath,String filePath){
 digesterRulesURL= getURL(rulePath);
 fileURL = getURL(filePath);
// digesterRulesURL= getClass().getClassLoader().getResource(rulePath);
 }
 public RuleLoader(URL digesterRulesURL,URL fileURL){
 this.digesterRulesURL = digesterRulesURL;
 this.fileURL = fileURL;
 }

 public static RuleLoader getXmlRuleLoader(String filePath){
 URL url = getURL("classpath:com/aisino/common/parse/xml-rules.xml");
 return new RuleLoader(url,getURL(filePath));
 }

 /**
 *
 * 自定义指定规则 
 * 需要对Digester 解析规则熟悉
 * @return
 */
 public List parseRules(){
 ClassLoader classLoader = getClass().getClassLoader();
 Object root = new ArrayList();
 try {
 DigesterLoader.load(digesterRulesURL, classLoader, fileURL,root);
 } catch (IOException e) {
 logger.error("IOException");
 } catch (SAXException e) {
 logger.error("SAX error");
 } catch (DigesterLoadingException e) {
 logger.error("an error occurs while parsing XML into Digester rules");
 }
 return (List)root;
 }

 /**
 *
 * 解析XML数据，并将数据保存至Map
 * 最多支持5级
 * XML节点:主要表示root、level1,level2,level3,level4,level5 
 * for example: 
 * <root nodeName="dataExchangePackage"> 
 * <level1 nodeName="envelopeInfo"> 
 * <level2 nodeName="sourceID" /> 
 * <level2 nodeName="destinationID" /> 
 * <level2 nodeName="destinationAppID" /> 
 * <level2 nodeName="businessType" /> 
 * <level2 nodeName="globalBusinessID" /> 
 * </level1>
 * <level1>
 * ...
 * </level1>
 </root>
 
 约定 ${}为节点对应nodeName值
 ${root} = dataExchangePackage
 ${level1}==envelopeInfo
 ${level2}==sourceID...
 * @param input
 * @return Map
 * 从map 获取值
 * 如果key 是叶子节点：则返回string,反之,为Map
 * 如果想取sourceID的值，则key=/dataExchangePackage/envelopeInfo/sourceID
 */
 public static Map reParseRules(String filePath,InputStream input){
 List rules = RuleLoader.getXmlRuleLoader(filePath).parseRules();
 if(rules != null && rules.size()>0){
 Digester digester = new Digester();
 Node root=(Node)rules.get(0);//根节点
 MapSetNextRule rule = new MapSetNextRule("put");
 addRule2Dister(digester, root,"", rule,true);
 try {
 Map valueMap = new HashMap();
 Map parseMap = (Map)digester.parse(input);
 valueMap.putAll(parseMap);
 afterRootMap(parseMap,valueMap,"");
 return valueMap;
 } catch (IOException e) {
 e.printStackTrace();
 } catch (SAXException e) {
 e.printStackTrace();
 }finally{
 digester=null;
 }
 }
 return null;
 }

 private static void afterRootMap(Map valueMap,Map destMap,String pattern){
 String fullPattern="";

 Iterator keys = valueMap.keySet().iterator();
 while(keys.hasNext()){
 Object key = keys.next();
 Object v = valueMap.get(key);
 fullPattern= pattern+"/"+key;
 if(v instanceof Map){
 afterRootMap((Map)v,destMap,fullPattern);
 }else{
 logger.debug(fullPattern+">>>>对应元素>"+v+" 放入返回栈中");
 destMap.put(fullPattern, v);
 }
 }
 }

 private static URL getURL(String resourceLocation){
 try {
 if (resourceLocation.startsWith(ResourceHelper.CLASSPATH_URL_PREFIX)) {
 return ResourceHelper.getURL(resourceLocation);
 } else if(resourceLocation.startsWith(ResourceHelper.FILE_URL_PREFIX)) {
 resourceLocation = StringUtils.replace(resourceLocation, ResourceHelper.FILE_URL_PREFIX, "");
 return new File(resourceLocation).toURI().toURL();
 }
 } catch (Exception e) {
 logger.error("解析XML路径时,出错");
 }
 return null;
 }


/**
* 递归添加解析规则
* @param digester
* @param node ：当前节点
* @param pattern:规则
* @param rule:支持map添加
* @param isRoot 是否为根节点
*/
private static void addRule2Dister(Digester digester,Node node,String pattern,MapSetNextRule rule,boolean isRoot){
 String fullPattern="";
 if(StringUtils.isNotBlank(pattern)){
 fullPattern = pattern+"/"+node.getNodeName();
 }else{
 fullPattern = node.getNodeName();
 }
 if(node.getChildren().size()>0){
 logger.debug(" add rules >>> digester.addObjectCreate("+fullPattern+", HashMap.class);");
 digester.addObjectCreate(fullPattern, HashMap.class);
 if(StringUtils.isNotBlank(node.getAttrs())){
 String[] attrs =StringUtils.split(node.getAttrs(),",");
 logger.debug(fullPattern+"有属性:"+ToStringBuilder.reflectionToString(attrs));
 for(int i=0;i<attrs.length;i++){
 String attr= attrs[i];
 logger.debug(" add rules >>> digester.addCallMethod("+fullPattern+",\"put\", 2)");
 logger.debug(" add rules >>> digester.addObjectParam("+fullPattern+",0, "+attr+")");
 logger.debug(" add rules >>> digester.addCallParam("+fullPattern+",1, "+attr+")");
 digester.addCallMethod(fullPattern, "put", 2);
 digester.addObjectParam(fullPattern, 0, attr);
 digester.addCallParam(fullPattern, 1,attr);
 }
 }
 if(!isRoot){//不是根节点
 logger.debug(" add rules >>> digester.addRule("+fullPattern+", rule);");
 digester.addRule(fullPattern, rule);
 }
 for(int i=0;i<node.getChildren().size();i++){
 Node child = (Node)node.getChildren().get(i);
 addRule2Dister(digester,child,fullPattern,rule,false);
 }

 }else{
 //叶子节点
 logger.debug("add rules >>> digester.addCallMethod("+fullPattern+", \"put\", 2)");
 digester.addCallMethod(fullPattern, "put", 2);
 logger.debug("add rules >>> digester.addObjectParam("+fullPattern+",0, "+node.getNodeName()+")");
 digester.addObjectParam(fullPattern, 0, node.getNodeName());
 logger.debug("add rules >>> digester.addCallParam("+fullPattern+",1)");
 digester.addCallParam(fullPattern, 1);
 }
 }
...
}

XML阅读器

public class XmlReader {
  private static final Log logger = LogFactory.getLog(XmlReader.class);
  private String rulePath;
  private String filePath;
  private RuleLoader rule;
  public XmlReader(String rulePath){
    this.rulePath = rulePath;
  }
  public List read(String filePath){
        rule = new RuleLoader(this.rulePath,filePath);
        return rule.parseRules();
  }
...
}

测试类

     public static void main(String args[]) throws Exception{
      InputStream resourceAsStream = Thread.currentThread().getContextClassLoader().getResourceAsStream("com/aisino/common/parse/101R.xml");
      Map m = RuleLoader.reParseRules("classpath:com/aisino/common/parse/xml-value-example.xml",resourceAsStream);
      System.out.println(m.get("/dataExchangePackage/envelopeInfo/destinationID"));
      System.out.println(m.get("/dataExchangePackage/transferInfo/age"));
      System.out.println(m.get("/dataExchangePackage/transferInfo/sex"));
        }

测试规则文件（定义XML解析规则）

暂时只支持5级。当然可以无限增加至6级别,8级...

Digester 好想不支持通配符吧。在这一点确实没有做到充分的可扩展。但是5级相当可以满足应用了

<?xml version="1.0"?>
<root nodeName="dataExchangePackage">
<level1 nodeName="envelopeInfo">
 <level2 nodeName="sourceID" />
 <level2 nodeName="destinationID" />
 <level2 nodeName="destinationAppID" />
 <level2 nodeName="businessType" />
 <level2 nodeName="globalBusinessID" />
</level1>
<level1 nodeName="transferInfo" attrs="age,sex">
 <level2 nodeName="senderID" />
 <level2 nodeName="receiverID" />
 <level2 nodeName="isRetry" />
 <level2 nodeName="sendTime" />
 <level2 nodeName="messageID" />
 <level2 nodeName="sourceMessageID" />
</level1>
<level1 nodeName="contentControl">
 <level2 nodeName="zip">
 <level3 nodeName="isZip"/>
 </level2>
 <level2 nodeName="encrypt">
 <level3 nodeName="isEncrypt"/>
 </level2>
 <level2 nodeName="code">
 <level3 nodeName="isCode"/>
 </level2>
</level1>
<level1 nodeName="packageInfo">
 <level2 nodeName="subPackage">
 <level3 nodeName="sequence"/>
 <level3 nodeName="content">
 <level4 nodeName="dj_nsrxx">
 <level5 nodeName="nsrsbh"/>
 <level5 nodeName="swjg_dm"/>
 <level5 nodeName="nsrdzdah"/>
 <level5 nodeName="nsrmc"/>
 <level5 nodeName="zjhm"/>
 <level5 nodeName="scjydz"/>
 <level5 nodeName="bsrmc"/>
 <level5 nodeName="dhhm"/>
 <level5 nodeName="djzclx_dm"/>
 <level5 nodeName="nsrzt_dm"/>
 <level5 nodeName="nsr_swjg_dm"/>
 </level4>
 </level3>
 </level2>
</level1>
<level1 nodeName="returnState">
 <level2 nodeName="returnCode"></level2>
 <level2 nodeName="returnMessageID"></level2>
 <level2 nodeName="returnMessage"></level2>
</level1>
</root>

测试数据

<?xml version="1.0" encoding="GBK"?>
<dataExchangePackage xmlns="http://www.chinatax.gov.cn/tirip/dataspec" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.chinatax.gov.cn/tirip/dataspec/dataExchangePackage.xsd" version="SW5001">
 <envelopeInfo>
 <sourceID>SJ</sourceID>
 <destinationID>YD</destinationID>
 <destinationAppID>SJQZ</destinationAppID>
 <businessType>WLFP101</businessType>
 <globalBusinessID>SJWLFP1012012011100184241</globalBusinessID>
 </envelopeInfo>
 <transferInfo age="88777" sex="男">
 <senderID>SJ</senderID>
 <receiverID>YD</receiverID>
 <isRetry/>
 <sendTime>2012-01-11 13:14:57:759</sendTime>
 <messageID>SJWLFP1012012011100184241</messageID>
 <sourceMessageID/>
 </transferInfo>
 <contentControl>
 <zip>
 <isZip>false</isZip>
 </zip>
 <encrypt>
 <isEncrypt>false</isEncrypt>
 </encrypt>
 <code>
 <isCode>false</isCode>
 </code>
 </contentControl>
 <packageInfo>
 <subPackage>
 <sequence>1</sequence>
 <content>
 <dj_nsrxx>
 <nsrsbh>350583729702365</nsrsbh>
 <swjg_dm>13505833100</swjg_dm><xgrq/>
 <nsrdzdah>350502001008324</nsrdzdah>
 <nsrmc>南安市水头康利石材有限公司</nsrmc>
 <zjhm>130302610511351</zjhm>
 <scjydz>南安市水头镇西锦村</scjydz>
 <bsrmc>陈姿颖</bsrmc>
 <dhhm>6981988</dhhm>
 <djzclx_dm>230</djzclx_dm>
 <nsrzt_dm>21</nsrzt_dm>
 <nsr_swjg_dm>13505833100</nsr_swjg_dm>
 </dj_nsrxx>
 </content>
 </subPackage>
 </packageInfo>
 <returnState>
 <returnCode>00</returnCode>
 <returnMessageID>YDWLFP1012012011100000001</returnMessageID>
 <returnMessage>成功</returnMessage>
 </returnState>
</dataExchangePackage>

测试结果

/dataExchangePackage/envelopeInfo/destinationID>>>>YD
/dataExchangePackage/transferInfo/age>>>88777
/dataExchangePackage/transferInfo/age>>>男