GBK格式文本转为UTF-8格式

左手的ㄟ右手 2023-08-17 17:05 82阅读 0赞

/**
     * gbk格式文本转为utf-8
     */
    public class GBKToUTF8Util {
        private static String path1 = "C:\\Users\\Desktop\\temp1\\";
        private static String path2 = "C:\\Users\\Desktop\\temp2\\";
    
        public static void main(String[] args) throws IOException {
            BufferedReader reader = null;
            BufferedWriter writer = null;
            try {
                File file1 = new File(path1 + "Test.java");
                File file2 = new File(path2 + "Test.java");
                if (!file2.exists()) {
                    file2.createNewFile();
                }
                reader = new BufferedReader(new InputStreamReader(new FileInputStream(file1), Charset.forName("GBK")));
                writer = new BufferedWriter(new OutputStreamWriter(new FileOutputStream(file2), Charset.forName("UTF-8")));
                String str;
                while ((str = reader.readLine()) != null) {
                    writer.write(str);
                    writer.newLine();
                }
            } catch (Exception e) {
                e.printStackTrace();
            } finally {
                try {
                    if (reader != null) {
                        reader.close();
                    }
                } catch (Exception e) {
                    e.printStackTrace();
                }
                try {
                    if (writer != null) {
                        writer.close();
                    }
                } catch (Exception e) {
                    e.printStackTrace();
                }
            }
        }
    
        /**
         * gbk转utf-8
         *
         * @param gbkStr
         * @return
         */
        public static String getUTF8StringFromGBKString(String gbkStr) {
            try {
                return new String(getUTF8BytesFromGBKString(gbkStr), "UTF-8");
            } catch (UnsupportedEncodingException e) {
                e.printStackTrace();
            }
            return null;
        }
    
        private static byte[] getUTF8BytesFromGBKString(String gbkStr) {
            int n = gbkStr.length();
            byte[] utfBytes = new byte[3 * n];
            int k = 0;
            for (int i = 0; i < n; i++) {
                int m = gbkStr.charAt(i);
                if (m < 128 && m >= 0) {
                    utfBytes[k++] = (byte) m;
                    continue;
                }
                utfBytes[k++] = (byte) (0xe0 | (m >> 12));
                utfBytes[k++] = (byte) (0x80 | ((m >> 6) & 0x3f));
                utfBytes[k++] = (byte) (0x80 | (m & 0x3f));
            }
            if (k < utfBytes.length) {
                byte[] tmp = new byte[k];
                System.arraycopy(utfBytes, 0, tmp, 0, k);
                return tmp;
            }
            return utfBytes;
        }
    }