On 01/23/2012 05:11 AM, Ulf Zibis wrote:
Am 23.01.2012 11:33, schrieb Alan Bateman:
On 23/01/2012 08:23, Joe Darcy wrote:
With these changes, the javadoc builds runs without additional
warnings and a specdiff against a reference copy of the javadoc
shows no unexpected changes.
I'm wondering because of </code>...</code> error, Alan pointed out.
In java.io.DataInput I see that the javadoc doesn't go beyond about
column 40 in many cases and maybe it would be good to clean this up
while you are there.
+1
It would be good, to have a place at jdk repo for such cosmetic
scripts (preferably for Unix + Windows). anyone could contribute for
enhancement + check the result at areas, one is working on, e.g. to
avoid unwanted changes like intentional line breaks outside the column
80 rule.
Well, there is the make/scripts directory in the the top-level JDK
repository forest.
Joe, thanks for catching my change proposal.
Would like to see me as contributor :-)
I'm wondering, that you only found 14 classes to change.
IIRC, there are at least usages of old <code>...</code> tags in
java.lang.Character + java.lang.AbstractStringBuilder.
I was not trying to replace all the <code></code> tags with {@code}!
Although others may attempt that later in JDK 8. (For the classes I
directly maintain, I long ago migrated to using {@code}.)
I was only attempting to replace the awkward HTML with <code></code>
used to render "\u..." with less awkward javadoc just using {@code}, but
if I was in a file, I replaced the other <code></code> usages to.
Cheers,
-Joe
