Re: [PATCH] Implement rehashing for associative arrays (Re: speeding up hash_search?)

[George Jones]

No problem.   Glad you fixed it.   It's been a LONG time since I've
actually written C, so probably best if someone current did it.

On the parameters, I suggest you consider exposing the at user level as a
switch or env var.   My use case was pathologically large (and would have
been better on, e.g. Spark if that were and option in the environment).
 Even the old behavior was probably good enough for most people.  It's only
when you start abusing bash to to "big data" that the problem shows up.

Thanks again, and impressive quick work !

---george jones



On Mon, Apr 20, 2020, 10:48 AM Koichi Murase <myoga.murase@gmail.com> wrote:

> 2020-04-20 23:05 Chet Ramey <chet.ramey@case.edu>:
> > On 4/20/20 8:49 AM, Greg Wooledge wrote:
> > > On Mon, Apr 20, 2020 at 06:48:44PM +0900, Koichi Murase wrote:
> > >> Also, I am sorry that I disturbed your plan for contributing to Bash.
> > >> I actually initially doubted that the insertion with the current
> > >> implementation is O(N), so I created the test first and then found
> > >> that it is an easy fix rather than reimplementing it by B-tree or
> > >> other data structures.  I couldn't stop my interest in how much it is
> > >> improved by the easy fix.
> > >
> > > This should in no way make the OP feel that they didn't contribute.
> > > Spotting and diagnosing problems is important work, even if their
> > > proposed patch wasn't selected as the best solution.
> >
> > This is quite true.
>
> Dear All,
>
> Yes, thank you for clarification and sorry for confusing writing.  I
> did not mean George's contribution disappeared, but just I have
> disturbed his five-step plan on his blog where the remaining four
> steps are now canceled.  I have to add that identifying the problem is
> definitely the most non-trivial part because it is relatively
> straightforward to fix the problem once the problem is identified :).
> Also, I would like to thank George for additional testing of my patch.
>
> Of course, this also applies to my patches.  I think I have sent more
> than ten patches so far and many of them have been applied, but I am
> anyway happy if the problems are solved.  You can always reject my
> patches when you have better solutions.
>
> --
> Koichi
>