[mkgmap-dev] highway=track with no road_class and road_speed set breaks routing ## was ## "No roads near target" bug in Schwabmünchen

Thu Apr 18 09:43:36 BST 2013

Hi,

I noticed that type 0x00 is a valid line type, so it's not a good idea to use this as a flag for "delete the line".
Also I was not able to get the unit tests running when the --list-styles option performs the checks.
Therefore I've added a new option --check-styles.

Attached is v6 of the patch. The binary is 

If you want to delete the unconnected road, use 
 add mkgmap:set_unconnected_type = none
instead of 
 add mkgmap:set_unconnected_type = 0x00

I found no easy way to implement a check that warns the user when he uses a routable type here, so this
is done only at runtime and will produce multiple messages like this:
SCHWERWIEGEND (StyledConverter): e:\bayern\63240001.o5m: type value in mkgmap:set_unconnected_type should not be a routable type: 0x02

and the corresponding way is ignored. I hope this makes sense?

Gerd

Date: Tue, 16 Apr 2013 08:48:26 +0200
From: franco.bez at web.de
To: mkgmap-dev at lists.mkgmap.org.uk
Subject: Re: [mkgmap-dev] highway=track with no road_class and road_speed set breaks routing ## was ## "No roads near target" bug in Schwabmünchen

    Hi Gerd,

    now I get both the warning during --list-styles and the Error during
    build. :-)

    when the offending line in the style is fixed, both the warning and
    the error disappear. :-)

    Looks good to me.

    Thanks a lot.

    Ciao,

       Franco

    Am 16.04.2013 07:01, schrieb Gerd
      Petermann:

      Hi Franco,

        > Subject: Re: [mkgmap-dev] highway=track with no
          road_class and road_speed set breaks routing ## was ## "No
          roads near target" bug in Schwabmünchen

          > 

          > franco_bez wrote

          > > 

          > > franco_bez wrote

          > >> I tested your binary with the proposed change on
          the Style, and now I

          > >> still get precisely 1 Error Message in my test
          case. Before there were

          > >> plenty (maybe 100) of them.

          The error message is only printed once now, before it was
          repeated for each case.

          > > The message confuses me its:

          > >> SEVERE (MapBuilder):

          > >>
          /home/franco/map_build/test_cases/tiles/63360004.o5m:
          Non-routable way

          > >> with routable type 0xf is used for a routable
          map. This leads to routing

          > >> errors. Use --list-styles to check the style.

          > > In all the style files I find no 0xf at all.

          > > Where ever this may come from ?

          > 

          > Could it be this line ?

          > highway=track & tracktype=grade2 [0x0f resolution 24]

          yes, 0xf is equal to 0x0f.  

          > 

          > but why does --list-styles not warn me ?

          Because I made another stupid mistake :-( 

          Sorry for that, I disabled one check while testing and forgot
          to enable it again.

          I'll try to code some unit tests for that when everybody is
          satisfied with the results.

          Attached is v5 of the patch and the binary is here:

          http://files.mkgmap.org.uk/download/109/mkgmap.jar

            Gerd

_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mkgmap.org.uk/pipermail/mkgmap-dev/attachments/20130418/f025fe39/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: routable_types_v6.patch
Type: application/octet-stream
Size: 20547 bytes
Desc: not available
Url : http://lists.mkgmap.org.uk/pipermail/mkgmap-dev/attachments/20130418/f025fe39/attachment-0001.obj 