RFR 8019484 Sync j.u.c.ConcurrentHashMap from 166 to tl
paul.sandoz at oracle.com
Mon Jul 1 16:19:39 UTC 2013
On Jul 1, 2013, at 3:19 PM, Doug Lea <dl at cs.oswego.edu> wrote:
> On 07/01/13 08:44, Paul Sandoz wrote:
>> The following is the final j.u.c patch that syncs CMH classes from 166 j.u.c to tl.
> Thanks for doing this!
>> ConcurrentMap.replaceAll in tl was more up to date than that in 166.
> Yes, we knew this would happen a few weeks ago when Mike committed,
> but I was still tracking lambda versions. Hopefully these mismatches
> will soon be things of the past.
OK. IIRC Mike wanted to take another pass at this stuff, so i went for the basic merge option.
>> Quite tricky to review CHM since there is a fair bit of code shuffling due to:
>> 480 * This file is organized to make things a little easier to follow
>> 481 * while reading than they might otherwise: First the main static
>> 482 * declarations and utilities, then fields, then main public
>> 483 * methods (with a few factorings of multiple public methods into
>> 484 * internal ones), then sizing methods, trees, traversers, and
>> 485 * bulk operations.
> Yes, sorry for the impact on diffs, but the old file organization
> (based on the original version of CHM) was getting too chaotic,
> so I rearranged in a way that better reflects structure.
I think the new arrangement is much better. JPRT results are good.
FWIW I found it easier just to linearly read though the new code, and contrast with searches within the old code.
More information about the core-libs-dev