在java中删除带有hadoop Wordcount的标点符号和HTML实体

import org.apache.commons.lang3.StringUtils;
import org.jsoup.Jsoup;
import org.jsoup.safety.Whitelist;

public static String removePunctuationAndHtmlEntities(String input) {
    // 删除标点符号
    String punctuationRemoved = StringUtils.remove(input, StringUtils.getPunctuation());

    // 删除HTML实体
    String htmlEntitiesRemoved = Jsoup.clean(punctuationRemoved, Whitelist.none());

    return htmlEntitiesRemoved;
}

public static void main(String[] args) {
    String input = "Hello, <b>world</b>!";

    String result = removePunctuationAndHtmlEntities(input);

    System.out.println(result);
}

Hello world