On 11/24/15 8:06 AM, Paul Benedict wrote:
*) List.of() says it returns "the newly created list" but it actually returns
the same empty list regardless. I don't think you want to imply a new list is
actually constructed for each call.
*) Map.of() same comment as above
*) Set.of() same comment as above
Right, see previous reply to Chris Hegarty.
*) Map.of(1/2/3/4/5/6/7/8/9) initializes the map to mystery capacities that
aren't explained. Should that be explained in a // comment or {@implNote} ?
It's not part of the spec, so no need for @implNote. But a // comment explaining
the numbers is probably warranted, since a couple people have asked about this.
*) Map.ofEntries(Entry<K,V>... entries) no need for check loop var "e" for
null since referencing it in the next line will throw NPE if null
I'll clean up.
*) As the size of the method parameters grow, it becomes more burdensome to
know which was the duplicate key in a Map/Set. Is it too expensive to check
after each element addition so that can be identified in the exception message
(the position at least)?
Good point. Since this isn't the "real" implementation I don't want to spend the
effort on this at the moment. My prototype compact set/map implementations add
the elements one at a time and report the first duplicate that they find. I
fully intend to work on the compact implementation, but on the off chance it
doesn't get in, the current implementation should be improved to do better
reporting in this case. I've filed a sub-task JDK-8143974 to track this.
s'marks
