Errata

The "in" variable is not defined.

Due to the bit shift operator having lower precedence than the addition operator, an extra pair of brackets is required. The line:

return (short) (((byte2 << 24) >>> 16) + (byte1 << 24) >>> 24);

should read:

return (short) (((byte2 << 24) >>> 16) + ((byte1 << 24) >>> 24));

Otherwise, the final >>> 24 operation is applied to byte2 as well as byte 1; this results in the most-significant byte read from the InputStream being discarded, as it suffers a net right-shift of 16 bits from its initial position at bits 0-7, rather than the required left-shift of 8 bits.

The signature of the method should not include an argument.

public byte readByte(int b) throws IOException

should read:

public byte readByte() throws IOException

The text should perhaps mention that (at least according to the Java 5/6 API) the methods getBytesRead() and getBytesWritten() are preferred over getTotalIn() and getTotalOut(), as they return a long rather than an int and thus can handle larger files.

The text should perhaps mention that (at least according to the Java 5/6 API) the methods getBytesRead() and getBytesWritten() are preferred over getTotalIn() and getTotalOut(), as they return a long rather than an int and thus can handle larger files.

In addition to the six setter methods mentioned in the text, there is a method setCompressedSize() for setting the size of the compressed entry data.

While the text states that the signatures for the two unpack() methods of the Pack200 class are

public void unpack(File in, OutputStream out) throws IOException
public void unpack(InputStream in, OutputStream out) throws IOException

they are in fact

public void unpack(File in, JarOutputStream out) throws IOException
public void unpack(InputStream in, JarOutputStream out) throws IOException

The third constructor, added in Java 1.4, has been omitted.

Some mistakes are in code fragment at page bottom:
[ERR] ByteBuffer buffer = buffer.allocate(8008);
[OK] ByteBuffer buffer = ByteBuffer.allocate(8008);

[ERR] roots.putDouble(i, Math.sqrt(i));
[OK] buffer.putDouble(i, Math.sqrt(i));

Some mistakes are in code fragment at page top:
[ERR] ByteBuffer buffer = buffer.allocate(1001 * DOUBLE_SIZE);
[OK] ByteBuffer buffer = ByteBuffer.allocate(1001 * DOUBLE_SIZE);

[ERR] roots.putDouble(i* DOUBLE_SIZE, Math.sqrt(i));
[OK] buffer.putDouble(i* DOUBLE_SIZE, Math.sqrt(i));

Sentence "Should the key become ready again in the future, is it included in the next
set returned by readyKeys()" is incorrect - there is no readyKeys(). readyKeys()
should be selectedKeys() (and really, as I understand it, it's more accurate to state
that it will be returned next next time select() is called.

"The File class provides two [createTempFile] methods *that create temporary files that exist only as long as the program runs*" (emphasis mine) -> as the author points out further down, there is a separate "deleteOnExit" method that must be called if you want temp files (or any other files) to be deleted on exit.

The FileTyper class will not print out files that consist of a single line, i.e. files with no line terminator. The following line should be added after the for-loop:

System.out.flush();