Base64 difference between C# and Java

Question

image is the string of an image file .

I have code as follows in C#:

Convert.ToBase64String(image);

and code as follows in Java: org.apache.commons.codec.binary.Base64.encodeBase64(image.getBytes())

The result is different.

Somebody says its because

• Java byte : -128 to 127

• C# byte : 0 to 255

But how can I fix this? How can I implement C#'s Convert.ToBase64String() in Java? I need the same result as in C# by using Java.

Answer 1

First you need to realise that a byte stores 256 values whether its signed or unsigned. If you want to get unsigned values from a signed byte (which is what Java supports) you can use & 0xFF

e.g.

byte[] bytes = { 0, 127, -128, -1};
for(byte b: bytes) {
  int unsigned = b & 0xFF;
  System.out.println(unsigned);
}

prints

0
127
128
255

The simple answer is you don't need a byte[] which has the same values. ;)

Answer 2

You're base64 encoding a string? What do you want that to do? You first need to convert the string to a sequence of bytes, choosing an encoding such as UTF-8 or UTF-16.

My guess is that you managed to use different encodings on both sides. Java's String.GetBytes() uses the default charset (Probably something like Latin1 on western windows versions). For C# you didn't post the relevant code.

To fix this, choose an encoding and use it explicitly on both sides. I recommend using UTF-8.
On the Java side you should use the correct method for encoding, so you don't end up with "modified UTF-8", but since I'm not a java programmer, I don't know which methods output modified UTF-8. I think it only happens if you abuse some internal serialization method.

signed vs. unsigned bytes should not be relevant here. The intermediate byte buffer will be different, but the original string, and the base64 string should be identical on both sides.