使用XML就应该简单点

对于现在越来越轻量级，越来越讲究速度和接近用户的应用来说，xml确实有点复杂了。

解析起来不仅耗内存，而且很复杂。

这就好像花了几千块钱买了个MS Office，但是80%的feature都用不着，还白白的耗着CPU和内存。

个人觉得，设置文件用XML其实挺好，因为设置文件一般并不太大，而且要求可读性强，还有很多乱七八糟的需求，可以利用XML的力量。

昨天搞chrome的设置，发现chrome的设置文件也是使用的json，读起来也是轻松愉快。

前阵子做了个程序，需要解析豆瓣API调用返回的XML。

真想说一句。

。

。

豆瓣你别用XML了。

。

。

至少，提供个json版的API调用吧。

（以上谨代表个人观点）

===================正文=================

解析豆瓣返回的xml，实在是不想用DOM这个重量级的玩意。

DOM这个玩意，说它强大好还是说它官僚好呢。

我倾向于使用SAXP解析。

但是现在面临的一个问题是，我需要根据xml节点的名字和属性值（一个或者多个）来决定当前的值是不是我想要的。

这就麻烦一点点。

第一反应是考虑xpath。

后来觉得不如自己做一个得了，权当是按需定制一个轻量级的xpath。

首先定义XMLSearchUnit类，这个类的实例用来描述一个需要在XML中搜索的值，值可以是xml节点的值，或者是节点的属性。

package com.deepnighttwo.resourceresolver.douban.resolver.utils;import java.util.HashMap;import java.util.Map;import org.xml.sax.Attributes;/*** * Represent a search task. Target could be value of a node or attribute of the* node.* * @author mzang*/publicclass XMLSearchUnit {// attribute values to be matched during searchprivate MapString, String> attributeMatchValidation =new HashMapString, String>();// if target is an attribute, then set this member to be the attribute name.// if it is null or empty, then means the target is node value.private String expectedAttr;// xml path, format is: /node_name/node_name/...private String xmlPath;public XMLSearchUnit(String xmlPath) {this.xmlPath = xmlPath;}/*** if current node meets the search conditions or not. Meets means the path* is correct and the attribute value is matched.* * @param path* @param attributes* @return*/publicboolean match(String path, Attributes attributes) {if (xmlPath.equals(path) ==false) {returnfalse;}for (String key : attributeMatchValidation.keySet()) {String exp = attributeMatchValidation.get(key);String compare = attributes.getValue(key);if (exp.equalsIgnoreCase(compare) ==false) {returnfalse;}}returntrue;}public MapString, String> getAttributeMatchValidation() {return attributeMatchValidation;}publicvoid addAttributeValidation(String key, String value) {attributeMatchValidation.put(key, value);}public String getXmlPath() {return xmlPath;}publicvoid setAttributeMatchValidation(MapString, String> attributeMatchValidation) {this.attributeMatchValidation = attributeMatchValidation;}public String getExpectedAttr() {return expectedAttr;}/*** if target is node value, then set expectedAttr to null. if target is an* attribute value, set it to be the attribute name.* * @param expectedAttr*/publicvoid setExpectedAttr(String expectedAttr) {this.expectedAttr = expectedAttr;}/*** hash code can be cached if all properties are not be be changed.*/@Overridepublicint hashCode() {finalint prime =31;int result =1;result = prime* result+ ((attributeMatchValidation ==null) ?0: attributeMatchValidation.hashCode());result = prime * result+ ((expectedAttr ==null) ?0 : expectedAttr.hashCode());result = prime * result + ((xmlPath ==null) ?0 : xmlPath.hashCode());return result;}@Overridepublicboolean equals(Object obj) {if (this== obj)returntrue;if (obj ==null)returnfalse;if (getClass() != obj.getClass())returnfalse;XMLSearchUnit other = (XMLSearchUnit) obj;if (attributeMatchValidation ==null) {if (other.attributeMatchValidation !=null)returnfalse;} elseif (!attributeMatchValidation.equals(other.attributeMatchValidation))returnfalse;if (expectedAttr ==null) {if (other.expectedAttr !=null)returnfalse;} elseif (!expectedAttr.equals(other.expectedAttr))returnfalse;if (xmlPath ==null) {if (other.xmlPath !=null)returnfalse;} elseif (!xmlPath.equals(other.xmlPath))returnfalse;returntrue;}}

这个类比较简单。

就是用一个hashmap保待匹配的attribut键值对，用一个字符串表示期待的attribute name，用一个字符串表示期待的node path。

然后就是如何在SAXP里用到这个类的实例去搜索了。

package com.deepnighttwo.resourceresolver.douban.resolver.utils;import java.io.InputStream;import java.util.ArrayList;import java.util.HashMap;import java.util.List;import java.util.Map;import javax.xml.parsers.SAXParser;import javax.xml.parsers.SAXParserFactory;import org.xml.sax.Attributes;import org.xml.sax.InputSource;import org.xml.sax.SAXException;import org.xml.sax.XMLReader;import org.xml.sax.helpers.DefaultHandler;/*** * SAXP parser working with XMLSearchUnit.* * @author mzang*/publicclass DoubanSearchParser extends DefaultHandler {// create and initial search unitspublicstaticfinal XMLSearchUnit DETAILS_LINK_API_PATH =new XMLSearchUnit("/feed/entry/id");publicstaticfinal XMLSearchUnit DETAILS_CONTENT_PATH =new XMLSearchUnit("/entry/summary");publicstaticfinal XMLSearchUnit DETAILS_TITLE_PATH =new XMLSearchUnit("/entry/title");publicstaticfinal XMLSearchUnit DETAILS_CHINESE_NAME_PATH =new XMLSearchUnit("/entry/db:attribute");publicstaticfinal XMLSearchUnit DETAILS_RATINGE_PATH =new XMLSearchUnit("/entry/gd:rating");publicstaticfinal XMLSearchUnit DETAILS_RATINGE_RATER_COUNT_PATH =new XMLSearchUnit("/entry/gd:rating");publicstaticfinal XMLSearchUnit DETAILS_LINK_URL_PATH =new XMLSearchUnit("/feed/entry/link");static {DETAILS_LINK_URL_PATH.addAttributeValidation("rel", "alternate");DETAILS_LINK_URL_PATH.setExpectedAttr("href");DETAILS_CHINESE_NAME_PATH.addAttributeValidation("lang", "zh_CN");DETAILS_CHINESE_NAME_PATH.addAttributeValidation("name", "aka");DETAILS_RATINGE_PATH.setExpectedAttr("average");DETAILS_RATINGE_RATER_COUNT_PATH.setExpectedAttr("numRaters");}// a map to store the XMLSearchUnit and valueprivate MapXMLSearchUnit, String> results =new HashMapXMLSearchUnit, String>();// a counter of search unit. if it is 0, then all search unit finds a match// value and the result of the XML will be skipped.privateint count =0;private StringBuilder path =new StringBuilder();privatestaticfinal String pathSeparater ="/";private XMLSearchUnit[] searchUnits;ListXMLSearchUnit> foundItems =new ArrayListXMLSearchUnit>();/*** constructor, accept XML input stream, 0 or more search unit instances.* * @param input* @param expectedPath* @return*/public MapXMLSearchUnit, String> parseResults(InputStream input,XMLSearchUnit... expectedPath) {for (XMLSearchUnit search : expectedPath) {results.put(search, null);}searchUnits = expectedPath;count = expectedPath.length;XMLReader xmlReader =null;try {SAXParserFactory spfactory = SAXParserFactory.newInstance();spfactory.setValidating(false);SAXParser saxParser = spfactory.newSAXParser();xmlReader = saxParser.getXMLReader();xmlReader.setContentHandler(this);xmlReader.parse(new InputSource(input));} catch (Exception e) {System.err.println(e);System.exit(1);}return results;}privatevoid addToPath(String addPath) {path.append(pathSeparater).append(addPath.toLowerCase());}privatevoid popPath() {int index = path.lastIndexOf(pathSeparater);// String removedPath = path.substring(index);path.delete(index, path.length());}@Overridepublicvoid startElement(String uri, String localName, String qName,Attributes attributes) throws SAXException {foundItems.clear();if (count ==0) {return;}// update pathaddToPath(qName);ListXMLSearchUnit> foundAttrItems =null;// check if current node matches search units. if it is a node value// search, then store it in a member variable named foundItems because// the value of the node is known only when reaches the end of the// node.but for attribute search, it value is known here. So then are// put in a local variable list named foundAttrItems.for (XMLSearchUnit unit : searchUnits) {if (unit.match(path.toString(), attributes) ==true) {if (unit.getExpectedAttr() ==null) {foundItems.add(unit);} else {if (foundAttrItems ==null) {foundAttrItems =new ArrayListXMLSearchUnit>();}foundAttrItems.add(unit);}}}// if no attribute match, return.if (foundAttrItems ==null) {return;}// fill search unit value using attribute value. update count.for (XMLSearchUnit attrUnit : foundAttrItems) {String attrValue = attributes.getValue(attrUnit.getExpectedAttr());if (results.get(attrUnit) ==null) {count--;}results.put(attrUnit, attrValue);count--;}}/*** if current node matches, the the node value is useful, store it.*/@Overridepublicvoid characters(char[] ch, int start, int length)throws SAXException {if (count ==0) {return;}if (foundItems.size() ==0) {return;}for (XMLSearchUnit unit : foundItems) {String content =new String(ch, start, length);if (results.get(unit) ==null) {count--;}results.put(unit, content);}}@Overridepublicvoid endElement(String uri, String localName, String qName)throws SAXException {foundItems.clear();if (count ==0) {return;}popPath();}}

使用,XML,就,应该,简,单点,对于,现在,越来越,