天天看点

java过滤html注释,java正则表达式过滤html标签

import java.util.regex.Matcher;

import java.util.regex.Pattern;

public classHtmlRegexpUtil {

privatefinalstaticStringregxpForHtml= "]*)>"; // 过滤所有以结尾的标签

privatefinalstaticStringregxpForImgTag= "]*)\\s*>"; // 找出IMG标签

privatefinalstaticStringregxpForImaTagSrcAttrib = "src=\"([^\"]+)\""; // 找出IMG标签的SRC属性

publicHtmlRegexpUtil() {

}

publicString replaceTag(String input) {

if (!hasSpecialChars(input)) {

return input;

}

StringBuffer filtered = newStringBuffer(input.length());

char c;

for (int i = 0; i <= input.length() - 1; i++){

c = input.charAt(i);

switch (c) {

case'

filtered.append("

break;

case'>':

filtered.append(">");

break;

case'"':

filtered.append(""");

break;

case'&':

filtered.append("&");

break;

default:

filtered.append(c);

}

}

return (filtered.toString());

}

publicbooleanhasSpecialChars(String input) {

boolean flag =false;

if ((input !=null) && (input.length() > 0)) {

char c;

for (int i = 0; i <= input.length() - 1; i++){

c = input.charAt(i);

switch (c) {

case'>':

flag = true;

break;

case'

flag = true;

break;

case'"':

flag = true;

break;

case'&':

flag = true;

break;

}

}

}

return flag;

}

publicstaticString filterHtml(String str) {

Pattern pattern = Pattern.compile(regxpForHtml);

Matcher matcher =pattern.matcher(str);

StringBuffer sb = new StringBuffer();

boolean result1 = matcher.find();

while (result1) {

matcher.appendReplacement(sb, "");

result1 = matcher.find();

}

matcher.appendTail(sb);

return sb.toString();

}

publicstaticString fiterHtmlTag(String str, String tag) {

String regxp = "]*)\\s*>";

Pattern pattern = Pattern.compile(regxp);

Matcher matcher =pattern.matcher(str);

StringBuffer sb = new StringBuffer();

boolean result1 = matcher.find();

while (result1) {

matcher.appendReplacement(sb, "");

result1 = matcher.find();

}

matcher.appendTail(sb);

return sb.toString();

}

publicstaticString replaceHtmlTag(String str, String beforeTag,

String tagAttrib, String startTag,String endTag) {

String regxpForTag = "]*)\\s*>";

String regxpForTagAttrib = tagAttrib + "=\"([^\"]+)\"";

Pattern patternForTag = Pattern.compile(regxpForTag);

Pattern patternForAttrib = Pattern.compile(regxpForTagAttrib);

Matcher matcherForTag =patternForTag.matcher(str);

StringBuffer sb = new StringBuffer();

boolean result = matcherForTag.find();

while (result) {

StringBuffer sbreplace = new StringBuffer();

Matcher matcherForAttrib =patternForAttrib.matcher(matcherForTag

.group(1));

if (matcherForAttrib.find()) {

matcherForAttrib.appendReplacement(sbreplace, startTag

+matcherForAttrib.group(1) + endTag);

}

matcherForTag.appendReplacement(sb,sbreplace.toString());

result = matcherForTag.find();

}

matcherForTag.appendTail(sb);

return sb.toString();

}

}