项目中用到的一个小工具类(字符过滤器)

字符串过滤，是比较常用的功能，我的当前项目也有用到。
如过滤User输入的Html,Js代码等，
由于过滤需求是可能变动的，
如客户又要你过滤一些脏词或者为内容中的url自动加上超链接等。
考虑“开-闭”(OCP)原则，
我决定使用装饰器(Decorator)模式。
首先定义Decorator接口:

package com.sanook.hompy.util.filter;

public interface StringFilter {

	public void setNextStringFilter(StringFilter stringFilter); //关联下一装饰器

	public String filter(String source); //处理过滤
}

然后用模板方法(Template Method)模式实现一个抽象的过滤器：
这样可以将相同的实现部分抽象出来。

package com.sanook.hompy.util.filter;

public abstract class AbstractStringFilter implements StringFilter {

	private StringFilter stringFilter;

	public void setNextStringFilter(StringFilter stringFilter) {
		this.stringFilter = stringFilter;
	}

	public String filter(String source) {
		String target = doFilter(source);
		if (stringFilter == null) {
			return target;
		}
		return stringFilter.filter(target);
	}

	// 模板抽象方法,传入要处理的string,返回处理完的string
	// 遵循模板方法doXXX命名方式
	public abstract String doFilter(String source); 

}

空的实现：

package com.sanook.hompy.util.filter;

public class EmptyFilter extends AbstractStringFilter {

	public String doFilter(String source) {
		return source;
	}

}

下面实现该接口的Html过滤：

package com.sanook.hompy.util.filter;

import org.apache.commons.lang.StringUtils;

public class HtmlFilter extends AbstractStringFilter {

	public String doFilter(String source) {
		source = StringUtils.replace(source, "<", "& lt;");
		source = StringUtils.replace(source, ">", "& gt;");
		source = StringUtils.replace(source, "&", "& amp;");
		source = StringUtils.replace(source, " ", "& nbsp;");
		source = StringUtils.replace(source, "\"", "& #0034;");
		source = StringUtils.replace(source, "\'", "& #0039;");
		return source;
	}

}

由于Decorator是嵌套结构(注：这里只用了前(before)装饰，所以看起来有点像链结构，如有必要，也可以加上后(after)装饰)，
它的调用关系需要组装，所以应该用建造者(Builder)模式或简单工厂模式。
这里使用简单工厂模式，工厂的获取用单例(Singleton)模式返回

package com.sanook.hompy.util.filter;

import java.util.HashMap;
import java.util.Map;

import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;

import com.sanook.hompy.util.manager.ConfigureManager;

public class StringFilterFactory {

	private static final Log log = LogFactory.getLog(StringFilterFactory.class);

	private static final StringFilterFactory stringFilterFactory = new StringFilterFactory();

	private Map filterMap = new HashMap();

	private StringFilterFactory() {
		//ConfigureManager是Hompy项目统一读取配置的类，以多例(Multi-Singleton)模式实现
		filterMap = ConfigureManager.getInstance("filter").getMap();
	}

	public static StringFilterFactory getInstance() {
		return stringFilterFactory;
	}

	//通过一个以逗号分割的过滤器引用名串，获取过滤器
	public StringFilter getStringFilterChain(String chain) {
		if (chain == null || chain.length() == 0) {
			return new EmptyFilter();
		}

		if ("all".equalsIgnoreCase(chain)) {
			return getAllStringFilterChain();
		}

		String[] filters = chain.split("\\,");
		return getStringFilterChain(filters);
	}

	public StringFilter getAllStringFilterChain() {
		String[] filters = (String[]) filterMap.values().toArray();
		return getStringFilterChain(filters);
	}

	public StringFilter getStringFilterChain(String[] filters) {
		if (filters == null || filters.length == 0) {
			return new EmptyFilter();
		}

		StringFilter[] stringFilters = new StringFilter[filters.length];
		for (int i = filters.length - 1; i >= 0; i--) {
			stringFilters[i] = getStringFilter(filters[i]);
			if (i != filters.length - 1) {
				stringFilters[i].setNextStringFilter(stringFilters[i + 1]);
			} else {
				stringFilters[i].setNextStringFilter(null);
			}
		}
		return stringFilters[0];
	}

	public StringFilter getStringFilter(String key) {
		if (key != null) {
			try {
				//通过类名反射得到过滤器的实例
				Class clazz = Class.forName((String) filterMap.get(key));
				StringFilter stringFilter = (StringFilter) clazz.newInstance();
				return stringFilter;
			} catch (ClassNotFoundException e) {
				e.printStackTrace();
				log.warn(e);
			} catch (InstantiationException e) {
				e.printStackTrace();
				log.warn(e);
			} catch (IllegalAccessException e) {
				e.printStackTrace();
				log.warn(e);
			}
		}
		return new EmptyFilter();
	}
}

配置文件filter.properties如下：

html=com.sanook.hompy.util.filter.HtmlFilter
url=com.sanook.hompy.util.filter.UrlFilter
js=com.sanook.hompy.util.filter.JavaScriptFilter
dirty=com.sanook.hompy.util.filter.DirtyWordFilter
quote=com.sanook.hompy.util.filter.QuotationMarkFilter
line=com.sanook.hompy.util.filter.NewLineFilter
lower=com.sanook.hompy.util.filter.LowerFilter

这些配置将通过ConfigureManager读到filterMap中，
其中key作为chain的引用名，value为过滤器对象名。

调用方式：

String chain = "html,js,dirty";
StringFilter stringFilter = StringFilterFactory.getInstance().getStringFilterChain(chain);
String source = "<b>aaaa</b>";
String result = stringFilter.filter(source);

现在如果你要扩展一个过滤器，只要继承AbstractStringFilter,实现doFilter(String source)方法，
在filter.properties加入其引用名即可。上面的配置示例中就是Hompy项目用到的一些过滤器。

Hompy项目以JSP作为View层，而StringFiler是属于展示逻辑，应由View层控制，所以，我使用了自定义标签。

package com.sanook.hompy.servlet.tag;

import com.sanook.hompy.util.filter.StringFilter;
import com.sanook.hompy.util.filter.StringFilterFactory;

public class FilterTag extends BodyOutTag {

	private static final long serialVersionUID = 1L;

	private String chain;
	
	public void setChain(String chain) {
		this.chain = chain;
	}

	public String doBody(String body) {
		StringFilter stringFilter = StringFilterFactory.getInstance()
				.getStringFilterChain(chain);
		return stringFilter.filter(body);
	}

}

其父类BodyOutTag是一个抽象类

package com.sanook.hompy.presentation.tag;

import javax.servlet.jsp.JspException;
import javax.servlet.jsp.JspTagException;
import javax.servlet.jsp.tagext.BodyTagSupport;

public abstract class BodyOutTag extends BodyTagSupport {
	
	private String body;
	
	public BodyOutTag() {
		super();
		init();
	}
	
	private void init() {
		body = null;
	}
	
	public void setBody(String body) {
		this.body = body;
	}
	
	public int doStartTag() throws JspException {
		return EVAL_BODY_BUFFERED;
	}
	
	public int doEndTag() throws JspException {
		if (body == null) {
			if (bodyContent != null && bodyContent.getString() != null) {
				body = bodyContent.getString().trim();
			} else {
				body = "";
			}
		}
		
		/*如果继承SimpleTagSupport，则用：
		if (body == null) {
			body = "";
			JspFragment body = getJspBody();
			if (body != null) {
				StringWriter writer = new StringWriter();
				body.invoke(writer);
				body = writer.toString();
			}
		}*/
		
		body = doBody(body);
		
		try {
			pageContext.getOut().print(body == null ? "" : body);
		} catch (java.io.IOException ex) {
			throw new JspTagException(ex.getMessage());
		}
		body = null;
		return EVAL_PAGE;
	}
	
	public void release() {
		super.release();
		init();
	}
	
	public abstract String doBody(String body);

}

配置/WEB-INF/tld/hompy-string.tld,
(我将其归纳在string处理namespace内)

<?xml version="1.0" encoding="UTF-8" ?>

<taglib xmlns="http://java.sun.com/xml/ns/j2ee" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://java.sun.com/xml/ns/j2ee http://java.sun.com/xml/ns/j2ee/web-jsptaglibrary_2_0.xsd" version="2.0">
	<description>hompy string tag library</description>
	<display-name>string</display-name>
	<tlib-version>1.0</tlib-version>
	<short-name>s</short-name>
	<uri>http://hompy.sanook.com/tag/string</uri>
	<tag>
		<description></description>
		<name>filter</name>
		<tag-class>com.sanook.hompy.servlet.tag.FilterTag</tag-class>
		<body-content>JSP</body-content>
		<attribute>
			<description>body</description>
			<name>body</name>
			<required>false</required>
			<rtexprvalue>true</rtexprvalue>
			<type>java.lang.String</type>
		</attribute>
		<attribute>
			<description>filter chain key, separator is ,</description>
			<name>chain</name>
			<required>true</required>
			<rtexprvalue>true</rtexprvalue>
		</attribute>
	</tag>
</taglib>

在web.xml的适当位置加入:

<taglib>
	<taglib-uri>hompy-string</taglib-uri>
	<taglib-location>/WEB-INF/tld/hompy-string.tld</taglib-location>
</taglib>

在jsp页面中使用如下:

<%@ page language="java" contentType="text/html" pageEncoding="UTF-8" isELIgnored="false"%>
<%@ taglib uri="hompy-string" prefix="s"%>
<html>
	<body>
		Test Filter: <s:filter chain="html,js,dirty">${picture.title}</s:filter>
	</body>
</html>

希望各位 帮忙重构。
Like Refactor!

Jive论坛中的过滤功能是采用的也是装饰器(Decorator)模式.可以参考