Hi Alan,
This one caught my eye:
src/share/classes/java/net/AbstractPlainDatagramSocketImpl.java
This entire class seems to re-declare inherited methods from
DatagramSocketImpl for no obvious reason and mess up the javadoc in the
process. In many/most cases there is no need to redeclare the methods
AFAICS. If there is some subtle reason then inheritance of the existing
javadoc should suffice in the majority of cases.
David
-----
On 30/09/2012 6:52 AM, Alan Bateman wrote:
A small number of javadoc warnings have sneaked into the jdk8 builds
lately. They are easily fixed but it's a good opportunity to try out
Jon's doccheck tool [1], built on the DocTree API, and get a better
workout. I decided to mostly focus on -Xmsgs:reference, which covers
references like @param and @throws. I fixed a small number of html
issues along the way but that was not my focus. The webrev, which fixes
up most of the reference issues in roughly the core area, is here:
http://cr.openjdk.java.net/~alanb/8000269/webrev/index.html
When I say "most" then I excluded j.u.c and java.lang.invoke.
The reason that so many files are changed is because doccheck looks at
private and package-private classes and methods that wouldn't normally
be in the javadoc. Also, no foray into the *Permission classes is
possible without fixing up at least a few formatting issues. Note that
there is one non-javadoc change in this webrev; in java.net.Inet4Address
I have removed a private static field that has not been used for several
releases.
The changes are trivial and easy to review. At it covers many areas then
partial reviews are okay too.
Thanks,
Alan.
[1]
http://mail.openjdk.java.net/pipermail/compiler-dev/2012-September/004800.html
