MessageDigest sha = MessageDigest.getInstance("SHA-1");
MessageDigest md5 = MessageDigest.getInstance("MD5");
InputStream input = ...;
byte buffer = new byte[BUFFER_SIZE];
while((len = input.read(buffer)) >= 0)
byte shaDigest = sha.digest();
byte md5Digest = md5.digest();
Jim, could you show me an example, please? Also, does this mean that each digest will be keeping in memory the read bytes from the InputStream?
No, each digest will keep only its own work area. You will be reading the data stream once.
Right... I see what you mean now. I have to update the digest upon every read.
The Javadoc is pretty clear. You can use clone only to calculate different intermediate digests using the same algorithm. You cannot use DigestInputStream to calculate different digest algorithms without reading the stream multiple times. You must use a regular InputStream and multiple MessageDigest objects; read the data once, passing each buffer to all MessageDigest objects to get multiple digests with different algorithms.
You could easily encapsulate this in your own variant of DigestInputStream, say MultipleDigestInputStream that follows the same general approach but accepts a collection of MessageDigest objects or algorithm names.