检测某个txt文件是gbk编码还是utf8编码

最新推荐文章于 2024-05-10 13:15:11 发布

sfshine

最新推荐文章于 2024-05-10 13:15:11 发布

阅读量1.8k

点赞数

分类专栏： Android

Android 专栏收录该内容

471 篇文章 0 订阅

订阅专栏

阅读器需要用到

public String getCharset() throws IOException {
        mBookFile = new RandomAccessFile(strFilePath, "r");
        byte[] bytes = new byte[200];
        mBookFile.read(bytes);
        mCharset = EncodeDetector.detect(bytes);      
        Log.e("TAG", "The charset is " + mCharset);
        return mCharset; 

    }

package com.example.novelreader;

public class EncodeDetector {

    public final static String UTF8 = "UTF-8"; 
    public final static String GBK = "GBK"; 
    /**
     * 检测
     * @param data
     * @return
     */
    public static String detect(byte[] data) {
        int utfCount = 0;
        boolean isUtf8 = true;
        for (int i = 0; i < data.length; i++) {
            byte b0 = data[i];
            if (utfCount == 0) {
                if ((b0 & 0x80) != 0) {
                    if ((b0 & 0xE0) == 0xC0) {
                        utfCount = 1;
                    } else if ((b0 & 0xF0) == 0xE0) {
                        utfCount = 2;
                    } else if ((b0 & 0xF8) == 0xF0) {
                        utfCount = 3;
                    } else if ((b0 & 0xFC) == 0xF8) {
                        utfCount = 4;
                    } else if ((b0 & 0xFE) == 0) {
                        utfCount = 5;
                    } else {
                        isUtf8 = false;
                        break;
                    }
                }
            } else {
                if ((b0 & 0xC0) != 0x80) {
                    isUtf8 = false;
                    break;
                }
                --utfCount;
            }
        }

        if (isUtf8) {
            return UTF8;
        } else {
            return GBK;
        }
    }
}