C#学习学习笔记的第四篇_验证txt的编码格式，解决读取文件乱码问题

在项目上遇到了读取txt文件乱码，使用的解码格式是Encoding.Default。使用gb2312也是乱码。

后查阅资料发现可以先判断txt文本文件的编码格式，然后再读取

            string fileName = @"C:\Users\xongwb\Desktop\ttt.txt";
            var tm=GetFileEncodeType(fileName);
            StreamReader reader = new StreamReader(fileName, tm);
            var msg = reader.ReadToEnd();
        

解析文件的编码格式方法

        public System.Text.Encoding GetFileEncodeType(string filename)
        {
            using (FileStream fs = new FileStream(filename, System.IO.FileMode.Open, System.IO.FileAccess.Read))
            {
                byte[] Unicode = new byte[] { 0xFF, 0xFE, 0x41 };
                byte[] UnicodeBIG = new byte[] { 0xFE, 0xFF, 0x00 };
                byte[] UTF8 = new byte[] { 0xEF, 0xBB, 0xBF };//带BOM
                Encoding reVal = Encoding.Default;
                BinaryReader br = new BinaryReader(fs);
                int length;
                int.TryParse(fs.Length.ToString(), out length);
                byte[] ss = br.ReadBytes(length);
                if (IsUTF8Bytes(ss) ||
                    (ss[0] == UTF8[0] && ss[1] == UTF8[1] && ss[2] == UTF8[2]))
                    reVal = Encoding.UTF8;
                else if (ss[0] == UnicodeBIG[0] && ss[1] == UnicodeBIG[1] && ss[2] == UnicodeBIG[2])
                    reVal = Encoding.BigEndianUnicode;
                else if (ss[0] == Unicode[0] && ss[1] == Unicode[1] && ss[2] == Unicode[2])
                    reVal = Encoding.Unicode;
                br.Close();

                return reVal;
            }

        }
        /// <summary>
        /// 判断是否是不带 BOM 的 UTF8 格式
        /// </summary>
        /// <param name=“data“></param>
        /// <returns></returns>
        private static bool IsUTF8Bytes(byte[] data)
        {
            int charByteCounter = 1;//计算当前正分析的字符应还有的字节数
            byte curByte;//当前分析的字节
            for (int i = 0; i < data.Length; i++)
            {
                curByte = data[i];
                if (charByteCounter == 1)
                {
                    if (curByte >= 0x80)
                    {
                        //判断当前
                        while (((curByte <<= 1) & 0x80) != 0)
                        {
                            charByteCounter++;
                        }
                        //标记位首位若为非0 则至少以2个1开始，如：110XXXXX.....1111110X
                        if (charByteCounter == 1 || charByteCounter > 6)
                        {
                            return false;
                        }
                    }
                }
                else
                {
                    //若是UTF-8 此时第一位必须为1
                    if ((curByte & 0xC0) != 0x80)
                    {
                        return false;
                    }
                    charByteCounter--;
                }
            }
            if (charByteCounter > 1)
            {
                throw new Exception("非预期的byte格式");
            }
            return true;
        }

    }