Checksums in Java

1. Overview

In this mini-article, we'll provide a brief explanation of what checksums are and show how to use some of Java's built-in features for calculating checksums.

2. Checksums and Common Algorithms

Essentially, a checksum is a minified representation of a binary stream of data.

Checksums are commonly used for network programming in order to check that a complete message has been received. Upon receiving a new message, the checksum can be recomputed and compared to the received checksum to ensure that no bits have been lost. Additionally, they may also be useful for file management, for instance, to compare files or to detect changes.

There are several common algorithms for creating checksums, such as Adler32 and CRC32. These algorithms work by converting a sequence of data or bytes into a much smaller sequence of letters and numbers.  They are designed such that any small change in the input will result in a vastly different calculated checksum.

Let's take a look at Java's support for CRC32. Note that while CRC32 may be useful for checksums, it's not recommended for secure operations, like hashing a password.

3. Checksum From a String or Byte Array

The first thing we need to do is to obtain the input to the checksum algorithm.

If we're starting with a String, we can use the getBytes() method to get a byte array from a String:

String test = "test";
byte[] bytes = test.getBytes();

Next, we can calculate the checksum using the byte array:

public static long getCRC32Checksum(byte[] bytes) {
    Checksum crc32 = new CRC32();
    crc32.update(bytes, 0, bytes.length);
    return crc32.getValue();
}

Here, we are using Java's built-in CRC32 class. Once the class is instantiated, we use the update method to update the Checksum instance with the bytes from the input.

Simply put, the update method replaces the bytes held by the CRC32 Object – this helps with code re-use and negates the need to create new instances of Checksum. The CRC32 class provides a few overridden methods to replace either the whole byte array or a few bytes within it.

Finally, after setting the bytes, we export the checksum with the getValue method.

4. Checksum From an InputStream

When dealing with larger data sets of binary data, the above approach would not be very memory-efficient as every byte is loaded into memory. When we have an InputStream, we may opt to use CheckedInputStream to create our checksum. By using this approach, we can define how many bytes are processed at any one time.

In this example, we process a given amount of bytes at the time until we reach the end of the stream. The checksum value is then available from the CheckedInputStream:

public static long getChecksumCRC32(InputStream stream, int bufferSize) 
  throws IOException {
    CheckedInputStream checkedInputStream = new CheckedInputStream(stream, new CRC32());
    byte[] buffer = new byte[bufferSize];
    while (checkedInputStream.read(buffer, 0, buffer.length) >= 0) {}
    return checkedInputStream.getChecksum().getValue();
}

5. Conclusion

In this tutorial, we look at how to generate checksums from byte arrays and InputStreams using Java's CRC32 support.

As always, the code is available over on GitHub.