jsp/java切断)の内容の方法と技巧
4488 ワード
http://www.iteye.com/topic/730396
package com.tools;
import java.util.ArrayList;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
@SuppressWarnings({"unchecked" })
public class CutHtml {
static String htmlMatch = "";
// html html
public static String removeMatchHtmlTag() {
// String html="<p></p><table><tr><td></td><td></td></tr></table>";
Pattern p = Pattern.compile("<([a-zA-Z]+)[^<>]*>(.*?)</\\1>");
Matcher m = p.matcher(htmlMatch);
if (m.find()) {
// System.out.println(htmlMatch);
htmlMatch = htmlMatch.replaceAll("<([a-zA-Z]+)[^<>]*>(.*?)</\\1>","$2");
removeMatchHtmlTag();
}
return htmlMatch;
}
public static String subStringHTML(String param, int length, String endWith) {
if(length<1) {System.out.println("length must >0");return null;}
if(param.length()<length){return param;}
StringBuffer result = new StringBuffer();
StringBuffer str = new StringBuffer();
int n = 0;
char temp;
boolean isCode = false; // HTML
boolean isHTML = false; // HTML ,
for (int i = 0; i < param.length(); i++) {
temp = param.charAt(i);
if (temp == ’<’) {
isCode = true;
}
else if (temp == ’&’) {
isHTML = true;
}
else if (temp == ’>’ && isCode) {
n = n - 1;
isCode = false;
}
else if (temp == ’;’ && isHTML) {
isHTML = false;
}
if (!isCode && !isHTML) {
n = n + 1;
// UNICODE
if ((temp + "").getBytes().length > 1) {
n = n + 1;
}
str.append(temp);
}
result.append(temp);
if (n >= length) {
break;
}
}
result.append(endWith);
// HTML
String temp_result = result.toString().replaceAll("(>)[^<>]*(<?)","$1$2");
// HTML
temp_result = temp_result
.replaceAll("<(AREA|BASE|BASEFONT|BODY|BR|COL|COLGROUP|DD|DT|FRAME|HEAD|HR|HTML|IMG|INPUT|ISINDEX|LI|LINK|META|OPTION|P|PARAM|TBODY|TD|TFOOT|TH|THEAD|TR|area|base|basefont|body|br|col|colgroup|dd|dt|frame|head|hr|html|img|input|isindex|li|link|meta|option|p|param|tbody|td|tfoot|th|thead|tr)[^<>]*/>","");
// HTML
// temp_result=temp_result.replaceAll("<([a-zA-Z]+)[^<>]*>(.*?)</\\1>","$2");
htmlMatch = temp_result;
temp_result = removeMatchHtmlTag();
//System.out.println("6666:" + temp_result);
//
Pattern p = Pattern.compile("<([a-zA-Z]+)[^<>]*>");
Matcher m = p.matcher(temp_result);
List endHTML = new ArrayList();
while (m.find()) {
endHTML.add(m.group(1));
}
// HTML
for (int i = endHTML.size() - 1; i >= 0; i--) {
result.append("</");
result.append(endHTML.get(i));
result.append(">");
}
return result.toString();
}
public static void main(String args[]) {
String param = "<html><p><p>0<img src=’xx.gif’/>http://www.dukai168.cn ©</p><table><tr><td>678901</td><td>21111111111</td></tr></table></p></html>";
System.out.println(CutHtml.subStringHTML(param,100, "..."));
}
}