Bug ID: JDK-6434084 (doc thread) New ThreadLocal SerialNum example still has problems

Type: Bug
Component: core-libs
Sub-Component: java.lang
Affected Version: 6,6u23,6u24

Priority: P3
Status: Resolved
Resolution: Fixed
OS: generic,linux,windows_7
CPU: generic,x86

Submitted: 2006-06-05
Updated: 2011-03-14
Resolved: 2006-07-01

JDK 6
6 b91Fixed

6202942 rfe 4 18M 1  Fix Delivered pete.soper java/lang (spec thread) ThreadLocal example should use generics and autoboxing

Comment added by    (###@###.###) : 
Wouldn't it be slightly better for the example to use getAndIncrement() rather than getAndAdd(1)? And a few finals wouldn't go amiss. And an @Override.

Oh and the HTML is slightly wrong, The  disappears, so the example makes even less sense than it did beforehand. {@code } would make source and docs readable.

EVALUATION Doc changes in place, exercise ("unit") test being developed.
24-06-2006
EVALUATION This will be ready for review soon.
06-06-2006
SUGGESTED FIX I have attached and copied below a short patch. I didn't use {@code } because the JavaDoc parser doesn't seem to be able to cope with the { @Override. It uses < and @ which is a bit ugly but works. I thought about wrapping the ThreadLocal line, but it looks fine as it is on the HTML page. j2se/src/share/classes/java/lang/ThreadLocal.java 2006-06-04 08:23:52.000000000 +0100 +++ java/lang/ThreadLocal.java 2006-06-06 10:50:33.000000000 +0100 @@ -25,23 +25,23 @@ * unchanged on subsequent calls.) * <pre> * import java.util.concurrent.atomic.AtomicInteger; + * * public class SerialNum { * * // The next serial number to be assigned. Starts at zero. - * private static AtomicInteger nextSerialNum = new AtomicInteger(); + * private static final AtomicInteger nextSerialNum = new AtomicInteger(); * - * private static ThreadLocal<Integer> serialNum = new ThreadLocal<Integer>() { - * + * private static final ThreadLocal<Integer> serialNum = new ThreadLocal<Integer>() { + * @Override * protected Integer initialValue() { - * return nextSerialNum.getAndAdd(1); + * return nextSerialNum.getAndIncrement(); * } * }; * * public static int get() { * return serialNum.get(); * } - * } - * </pre> + * }</pre> * * <p>Each thread holds an implicit reference to its copy of a thread-local * variable as long as the thread is alive and the <tt>ThreadLocal</tt> --- j2se/src/share/classes/java/lang/ThreadLocal.java 2006-06-04 08:23:52.000000000 +0100 +++ java/lang/ThreadLocal.java 2006-06-06 10:50:33.000000000 +0100 @@ -25,23 +25,23 @@ * unchanged on subsequent calls.) * <pre> * import java.util.concurrent.atomic.AtomicInteger; + * * public class SerialNum { * * // The next serial number to be assigned. Starts at zero. - * private static AtomicInteger nextSerialNum = new AtomicInteger(); + * private static final AtomicInteger nextSerialNum = new AtomicInteger(); * - * private static ThreadLocal<Integer> serialNum = new ThreadLocal<Integer>() { - * + * private static final ThreadLocal<Integer> serialNum = new ThreadLocal<Integer>() { + * @Override * protected Integer initialValue() { - * return nextSerialNum.getAndAdd(1); + * return nextSerialNum.getAndIncrement(); * } * }; * * public static int get() { * return serialNum.get(); * } - * } - * </pre> + * }</pre> * * <p>Each thread holds an implicit reference to its copy of a thread-local * variable as long as the thread is alive and the <tt>ThreadLocal</tt>
06-06-2006
EVALUATION These fixes all seem valid.
05-06-2006

Duplicate :	JDK-7020229 - ThreadLocal javadoc example is broken
Duplicate :	JDK-7027197 - Error in ThreadLocal javadoc example
Duplicate :	JDK-7017494 - Example in java.lang.ThreadLocal documentation is incorrect.
Relates :	JDK-6202942 - (spec thread) ThreadLocal example should use generics and autoboxing
Relates :	JDK-6446082 - new test/java/lang/ThreadLocal/TestUniqueThreadIdGenerator.java fails to compile