[mkgmap-dev] Commit r4314: fix assertion error when option --make-poi-index is used

Tue Oct 22 16:19:48 BST 2019

Hi Ticker,

thanks for review.
Collections.sort(list) simply calls list.sort(null), so performance should not matter.
I agree that the stream().ap() stuff doesn't inprove readability, I'll probably remove that from the patch.

Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap at jagit.co.uk>
Gesendet: Dienstag, 22. Oktober 2019 17:15
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Commit r4314: fix assertion error when option --make-poi-index is used

Hi Gerd

Looking at sort6.patch:

Generally
-               Collections.sort(keys);

+               keys.sort(null);
I think the obj.sort() form looks better. Possibly more efficient?

Considering the lambdas. In some cases isn't it clearer to put a
   int compareTo(T other) {...}
function in the definition of the records to be sorted that "implements
Comparable<T>"

I haven't yet worked out the meaning of constructs like:
    pil.stream().map(index -> sort.createSortKey(index,
                                       index.getName()))
   .sorted()
   .forEach(key -> key.getObject().write(writer));

On Tue, 2019-10-22 at 08:33 +0000, Gerd Petermann wrote:
> Hi Ticker,
>
> sorry, I think I thought that Steve should take care and it slipped
> my mind.
> Unfortunately Steve seems to have lost interest in reviewing patches.
> So yes, please review the patch and post a new version based on
> trunk.
>
> Gerd
> P.S.
> I just stumbled over this while checking a possible refactoring to
> change places where Collections.sort() is used.
> See attached patch which is work in progress.  I did not yet make up
> my mind regarding lambdas. They are nice to simplify code but
> sometimes it is much harder to debug them in Eclipse.
>
> ________________________________________
> Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag
> von Ticker Berkin <rwb-mkgmap at jagit.co.uk>
> Gesendet: Dienstag, 22. Oktober 2019 10:11
> An: mkgmap-dev at lists.mkgmap.org.uk
> Betreff: Re: [mkgmap-dev] Commit r4314: fix assertion error when
> option --make-poi-index is used
>
> Hi Gerd
>
> I noticed this problem in April and there was a fix to it in the
> patch that simplified/clarified point type/subtype and city range
> handling:
>
> http://www.mkgmap.org.uk/pipermail/mkgmap-dev/2019q2/029623.html
>
> The posting didn't elicit any responses, but I think the changes
> worthwhile. If you want to consider the patch, I'll need to check any
> interactions with more recent changes.
>
> Ticker
>
> On Mon, 2019-10-21 at 21:50 +0100, svn commit wrote:
> Version mkgmap-r4314 was committed by gerd on Mon, 21 Oct 2019
>
> fix assertion error when option --make-poi-index is used
> Regession from r4147 (img_io_3.patch)
>
> http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4314
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev