【Java】如何检测、替换4个字节的utf-8编码（此范围编码包含emoji表情）

> 参考的优秀文章

1、十分钟搞清字符集和字符编码

2、Java中byte与16进制字符串的互相转换

3、【异常处理】Incorrect string value: '\xF0\x90\x8D\x83...' for column... Emoji表情字符过滤的Java实现

4、Why a surrogate java regexp finds hypen-minus

> 如何检测、替换4个字节的utf-8编码（此范围编码包含emoji）

项目有个需求，是保存从手机端H5页面提交的信息。

大家知道，手机端输入法中经常有自带的表情，其中emoji表情非常流行，如果用户输入emoji表情，由于有部分emoji表情是4个字节的utf-8编码，我们的MySQL数据库在现有版本和编码设置下只能保存3个字节的utf-8编码（如要保存4个字节的utf-8编码则需升级版本和设置另一种编码）。相关信息可见文章《十分钟搞清字符集和字符编码》。

我们的需求不需要支持emoji表情，如果遇到emoji弹出提示或过滤即可。

通过浏览《【异常处理】Incorrect string value: '\xF0\x90\x8D\x83...' for column... Emoji表情字符过滤的Java实现》和《Why a surrogate java regexp finds hypen-minus》，我们得知通过以下代码进行替换：

msg.replaceAll("[\\ud800\\udc00-\\udbff\\udfff\\ud800-\\udfff]", "");

效果是OK的。

但是，由于能力原因，始终没能理解上述代码十六进制正则表达式的原理，自己写了一端代码来检测、替换4个字节的utf-8编码（但未能经过完整测试，仅用于描述大概思路）。

其中UTF-8编码规则阅读自《十分钟搞清字符集和字符编码》，字节与十六进制的转换参考自《Java中byte与16进制字符串的互相转换》。

package com.nicchagil.tc.emojifilter;

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

public class UTF8Utils {

    public static void main(String[] args) {
        String s = "琥珀蜜蜡由于硬度很低,打磨起来非123常简单,需要的工具也非常简单,自己去买蜜蜡是非常不划算的,完全可以自己磨蜜蜡原石的样子";
        System.out.println(UTF8Utils.bytesToHex(s.getBytes()));
        System.out.println(UTF8Utils.bytesToHex(UTF8Utils.remove4BytesUTF8Char(s)));
    }

    public static Map<String, Integer> hexMap = new HashMap<String, Integer>();
    public static Map<String, Integer> byteMap = new HashMap<String, Integer>();

    static {
        hexMap.put("0", 2);
        hexMap.put("1", 2);
        hexMap.put("2", 2);
        hexMap.put("3", 2);
        hexMap.put("4", 2);
        hexMap.put("5", 2);
        hexMap.put("6", 2);
        hexMap.put("7", 2);
        hexMap.put("c", 4);
        hexMap.put("d", 4);
        hexMap.put("e", 6);
        hexMap.put("f", 8);

        byteMap.put("0", 1);
        byteMap.put("1", 1);
        byteMap.put("2", 1);
        byteMap.put("3", 1);
        byteMap.put("4", 1);
        byteMap.put("5", 1);
        byteMap.put("6", 1);
        byteMap.put("7", 1);
        byteMap.put("c", 2);
        byteMap.put("d", 2);
        byteMap.put("e", 3);
        byteMap.put("f", 4);
    }

    /**
     * 是否包含4字节UTF-8编码的字符（先转换16进制再判断）
     * @param s 字符串
     * @return 是否包含4字节UTF-8编码的字符
     */
    public static boolean contains4BytesChar(String s) {
        if (s == null || s.trim().length() == 0) {
            return false;
        }

        String hex = UTF8Utils.bytesToHex(s.getBytes());
        System.out.println("full hex : " + hex);

        String firstChar = null;
        while (hex != null && hex.length() > 1) {
            firstChar = hex.substring(0, 1);
            System.out.println("firstChar : " + firstChar);

            if ("f".equals(firstChar)) {
                System.out.println("it is f start, it is 4 bytes, return.");
                return true;
            }

            if (hexMap.get(firstChar) == null) {
                System.out.println("it is f start, it is 4 bytes, return.");
                // todo, throw exception for this case
                return false;
            }

            hex = hex.substring(hexMap.get(firstChar), hex.length());
            System.out.println("remain hex : " + hex);
        }

        return false;
    }

    /**
     * 是否包含4字节UTF-8编码的字符
     * @param s 字符串
     * @return 是否包含4字节UTF-8编码的字符
     */
    public static boolean contains4BytesChar2(String s) {
        if (s == null || s.trim().length() == 0) {
            return false;
        }

        byte[] bytes = s.getBytes();

        if (bytes == null || bytes.length == 0) {
            return false;
        }

        int index = 0;
        byte b;
        String hex = null;
        String firstChar = null;
        int step;
        while (index <= bytes.length - 1) {
            System.out.println("while loop, index : " + index);
            b = bytes[index];

            hex = byteToHex(b);
            if (hex == null || hex.length() < 2) {
                System.out.println("fail to check whether contains 4 bytes char(1 byte hex char too short), default return false.");
                // todo, throw exception for this case
                return false;
            }

            firstChar = hex.substring(0, 1);

            if (firstChar.equals("f")) {
                return true;
            }

            if (byteMap.get(firstChar) == null) {
                System.out.println("fail to check whether contains 4 bytes char(no firstchar mapping), default return false.");
                // todo, throw exception for this case
                return false;
            }

            step = byteMap.get(firstChar);
            System.out.println("while loop, index : " + index + ", step : " + step);
            index = index + step;
        }

        return false;
    }

    /**
     * 去除4字节UTF-8编码的字符
     * @param s 字符串
     * @return 已去除4字节UTF-8编码的字符
     */
    public static byte[] remove4BytesUTF8Char(String s) {
        byte[] bytes = s.getBytes();
        byte[] removedBytes = new byte[bytes.length];
        int index = 0;

        String hex = null;
        String firstChar = null;
        for (int i = 0; i < bytes.length; ) {
            hex = UTF8Utils.byteToHex(bytes[i]);

            if (hex == null || hex.length() < 2) {
                System.out.println("fail to check whether contains 4 bytes char(1 byte hex char too short), default return false.");
                // todo, throw exception for this case
                return null;
            }

            firstChar = hex.substring(0, 1);

            if (byteMap.get(firstChar) == null) {
                System.out.println("fail to check whether contains 4 bytes char(no firstchar mapping), default return false.");
                // todo, throw exception for this case
                return null;
            }

            if (firstChar.equals("f")) {
                for (int j = 0; j < byteMap.get(firstChar); j++) {
                    i++;
                }
                continue;
            }

            for (int j = 0; j < byteMap.get(firstChar); j++) {
                removedBytes[index++] = bytes[i++];
            }
        }

        return Arrays.copyOfRange(removedBytes, 0, index);
    }

    /**
     * 将字符串的16进制转换为HEX，并按每个字符的16进制分隔格式化
     * @param s 字符串
     */
    public static String splitForReading(String s) {
        if (s == null || s.trim().length() == 0) {
            return "";
        }

        String hex = UTF8Utils.bytesToHex(s.getBytes());
        System.out.println("full hex : " + hex);

        if (hex == null || hex.length() == 0) {
            System.out.println("fail to translate the bytes to hex.");
            // todo, throw exception for this case
            return "";
        }

        StringBuilder sb = new StringBuilder();
        int index = 0;

        String firstChar = null;
        String splittedString = null;
        while (index < hex.length()) {
            firstChar = hex.substring(index, index + 1);

            if (hexMap.get(firstChar) == null) {
                System.out.println("fail to check whether contains 4 bytes char(no firstchar mapping), default return false.");
                // todo, throw exception for this case
                return "";
            }

            splittedString = hex.substring(index, index + hexMap.get(firstChar));
            sb.append(splittedString).append(" ");
            index = index + hexMap.get(firstChar);
        }

        System.out.println("formated sb : " + sb);
        return sb.toString();
    }

    /**
     * 字节数组转十六进制
     * @param bytes 字节数组
     * @return 十六进制
     */
    public static String bytesToHex(byte[] bytes) {
        if (bytes == null || bytes.length == 0) {
            return null;
        }

        StringBuilder sb = new StringBuilder();
        for (int i = 0; i < bytes.length; i++) {
            int r = bytes[i] & 0xFF;

            String hexResult = Integer.toHexString(r);
            if (hexResult.length() < 2) {
                sb.append(0); // 前补0
            }
            sb.append(hexResult);
        }

        return sb.toString();
    }

    /**
     * 字节转十六进制
     * @param b 字节
     * @return 十六进制
     */
    public static String byteToHex(byte b) {
        int r = b & 0xFF;
        String hexResult = Integer.toHexString(r);

        StringBuilder sb = new StringBuilder();
        if (hexResult.length() < 2) {
            sb.append(0); // 前补0
        }
        sb.append(hexResult);
        return sb.toString();
    }

}

在随便看下各种字符的UTF-8编码是什么：

package com.nicchagil.tc.emojifilter;

public class UTF8HexTester {

    public static void main(String[] args) {
        String s = "1";
        System.out.println("the hex of “" + s + "” : " + UTF8Utils.bytesToHex(s.getBytes()));

        s = "a";
        System.out.println("the hex of “" + s + "” : " + UTF8Utils.bytesToHex(s.getBytes()));

        s = "我";
        System.out.println("the hex of “" + s + "” : " + UTF8Utils.bytesToHex(s.getBytes()));

        s = "我很帅";
        System.out.println("the hex of “" + s + "” : " + UTF8Utils.bytesToHex(s.getBytes()));
    }

}

日志：

the hex of “1” : 31
the hex of “a” : 61
the hex of “我” : e68891
the hex of “我很帅” : e68891e5be88e5b885

> 搭建一个测试渠道来测试

由于emoji表情在PC不易输入，最好的输入途径始终在手机上，那么我们搭一个简单的web程序来接收emoji表情吧~

<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8">
<title>Emoji</title>
</head>
<script type="text/javascript" src="https://code.jquery.com/jquery-1.12.3.min.js"></script>
<body>

<form id="myform" action="http://192.168.1.3:8080/emoji/EmojiFilterServlet" >

    Input parameter :
    <input type='text' name='msg' />
    <br/>

    <input type='button' value=' ajax submit ' onclick="save();" />
    <input type='submit' value=' form submit ' />
</form>

</body>

<script type="text/javascript">

function save() {
    // alert('start save...');
    var data = $('#myform').serialize();
    // alert(data);

    $.ajax({
        type : "POST",
        url : "http://192.168.1.3:8080/emoji/EmojiFilterServlet",
        data : data,
        success : function(d) {
            alert(d);
        }
    });
}

</script>

</html>

package com.nicchagil.tc.emojifilter;

import java.io.IOException;
import java.nio.charset.Charset;

import javax.servlet.ServletException;
import javax.servlet.http.HttpServlet;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;

/**
 * Servlet implementation class EmojiFilterServlet
 */
public class EmojiFilterServlet extends HttpServlet {
    private static final long serialVersionUID = 1L;

    /**
     * @see HttpServlet#HttpServlet()
     */
    public EmojiFilterServlet() {
        super();
        // TODO Auto-generated constructor stub
    }

    /**
     * @see HttpServlet#doGet(HttpServletRequest request, HttpServletResponse response)
     */
    protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
        this.doPost(request, response);
    }

    /**
     * @see HttpServlet#doPost(HttpServletRequest request, HttpServletResponse response)
     */
    protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
        String msg = request.getParameter("msg");
        System.out.println("msg -> " + msg);
    }

}