java - 字节流转成字符串之后，在通过字符串转成字节流后的文件为什么会不一样？

Question

 public static void main(String[] args) throws Exception {
        File sourceFile = new File("/home/joy/图片/img1-lg.jpg");
        File tempFile = new File("/home/joy/桌面/TempFile");
        saveTempFile(sourceFile, tempFile);
        String str = byteToString(tempFile);
        File saveFile = new File("/home/joy/桌面/saveFile.jpg");
        StringTobyte(str,saveFile);
    }

    // 保存临时文件
    public static void saveTempFile(File sourceFile, File tempFile) throws Exception {
        FileInputStream inputStream = new FileInputStream(sourceFile);
        FileOutputStream fileOutputStream = new FileOutputStream(tempFile);
        byte[] buf = new byte[10 * 1024];
        int readLeng = 0;
        while ((readLeng = inputStream.read(buf)) != -1) {
            fileOutputStream.write(buf, 0, readLeng);
            fileOutputStream.flush();
        }
        fileOutputStream.close();
        inputStream.close();
    }

    // 文件字节转字符串
    public static String byteToString(File file) throws Exception {
        StringBuilder returnDatas = new StringBuilder();
        FileInputStream fileInputStream = new FileInputStream(file);
        byte[] buf = new byte[10 * 1024];
        int readLenth = 0;
        while ((readLenth = fileInputStream.read(buf)) != -1) {
            returnDatas.append(new String(buf,0,readLenth));
        }
        fileInputStream.close();
        return returnDatas.toString();
    }

    // 字符串转文件
    public static void StringTobyte(String str,File file) throws Exception {
        FileOutputStream fileOutputStream = new FileOutputStream(file);
        fileOutputStream.write(str.getBytes());
        fileOutputStream.close();
    }
    
    

代码大意：把一张图片通过字节流转成一个字符串，然后在获取该字符串的字节后还原为文件。
理解中，字符只是多带一层编码，还原成字节后应该还是能获取到源文件的呀。

Answer 1

        while ((readLenth = fileInputStream.read(buf)) != -1) {
            returnDatas.append(new String(buf,0,readLenth));
        }

二进制图片文件read出来的不是合法的字符串，其中无法解析的二进制数据会被滤掉
参考 javadoc 中的描述：

This method always replaces malformed-input and unmappable-character sequences with this charset's default replacement string.

Answer 2

编码和解码不一定是无损的。你从一个图片文件读出一个能无损编解码的字节流的可能性非常小。

new String( new byte[] { 0xff } ).getBytes()
=>  [-17, -65, -67]

Answer 3

首先保证序列化，你可以用base64，才能保证转回来没问题