- mkgmap-dev - The mkgmap lists

Dead end check
by Mike Baggaley 18 Feb '20

18 Feb '20

Hi all, I have set logging to enable warnings in StyledConverter, and have been getting a few unexpected results from the dead end check. It seems that using the default style and latest mkgmap release, if a way has mkgmap:dead-end-check=no set in OSM, this information is not loaded, so when the dead end check is run, ways with this tag are not omitted from the check. I have worked around this by adding a dummy line to the lines file: mkgmap:dead-end-check=no {set mkgmap:dead-end-check=no} The tag data is then loaded. It seems to me though that the code should include mkgmap:dead-end-check in the list of standard tags to be loaded, without needing anything specific in the style (or at least if dead end check is enabled). I would not expect OSM to have many of these, so it should not make any significant difference to the amount of memory needed by mkgmap or affect the execution time. An example is way 108322149 which is one of the ways where cars get on the train at Ashford. I suspect the same thing will happen with mkgmap:flare-check, mkgmap:dir-check and mkgmap:no-dir-check. Regards, Mike

3 7

errors in typ file
by Arndt Röhrig 17 Feb '20

17 Feb '20

4 10

best practices for parcel data?
by Greg Troxel 16 Feb '20

16 Feb '20

(Yes, I know parcel data does not belong in OSM. I am not putting it in, and not proposing to do so.) I am making maps for a my etrex 30 with a Geofabrik extract of Massachusetts in pbf, and files that I converted from MassGIS "parcels" data, which is basically shapefiles with tax lots and also easements. This results in polygons with tags variously as boundary=parcel boundary=easement I have the following diff, which results in a contour that is pretty visible and a very thin blue line that is "intermittent stream". The lot lines show as "contour line" and don't say depth on the Etrex 30. I am not firmly attached to this; I'm looking for a reasonable way to have lot lines and easements shown so that I can understand them. I suspect I could pick some not-really-used codepoints and put in TYP file content for them Other than TYP, does anyone have suggestions for a good way to do this? Is adding this to the default style sensible, because OSM doctrine is to merge parcel data when rendering? Or am I unusual and me carrying a patch is better? As always, thanks to everyone who has worked on mkgmap. My MassGIS shapefile to OSM conversion was some effort, but the mkgmap part was easy. Greg Index: resources/styles/default/lines =================================================================== --- resources/styles/default/lines (revision 4451) +++ resources/styles/default/lines (working copy) @@ -248,6 +248,10 @@ boundary=administrative [0x1c resolution 22] boundary=national [0x1e resolution 17] boundary=political [0x1c resolution 19] +# 0x23 is depth countour - thin. Wacky but useful. 0x1c is too heavy +boundary=parcel [0x23 resolution 20] +# wild guess +boundary=easement [0x26 resolution 20] barrier=wall | barrier=fence | barrier=hedge | barrier=ditch {add name='${barrier|subst:"_=> "}'} [0x17 resolution 24]

1 0

Re: [mkgmap-dev] Commit r4448: fix possible NPE introduced with r4432
by Mike Baggaley 15 Feb '20

15 Feb '20

Hi Gerd, as I understand it, exceptions are passed up the call stack until a handler is found. If no handler is found by the time main is reached, then the program crashes. There is no need for a re-throw. See https://www.geeksforgeeks.org/exceptions-in-java/ Cheers, Mike -----Original Message----- From: Gerd Petermann [mailto:gpetermann_muenchen@hotmail.com] Sent: 15 February 2020 12:34 To: Development list for mkgmap <mkgmap-dev(a)lists.mkgmap.org.uk> Subject: Re: [mkgmap-dev] Commit r4448: fix possible NPE introduced with r4432 Hi Mike, the re-throw is the difference. The exception is now passed to the caller. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces(a)lists.mkgmap.org.uk> im Auftrag von Mike Baggaley <mike.baggaley(a)basfordltc.co.uk> Gesendet: Donnerstag, 13. Februar 2020 11:09 An: 'Development list for mkgmap' Betreff: Re: [mkgmap-dev] Commit r4448: fix possible NPE introduced with r4432 >Version mkgmap-r4448 was committed by gerd on Thu, 13 Feb 2020 > >fix possible NPE introduced with r4432 >FileNotFoundException was ignored. May result in a NullPointerException when input file is not in the expected format. e.g. with >java -jar mkgmap.jar . Hi Gerd, I can't see how this update can make any difference. The change catches the FileNotFoundException and immediately re-throws it, so it effectively does nothing. Looking at the call hierarchy, FileNotFoundException is caught in createRawData, makeMap and loadCoastlines so all the functions that call the load function directly or indirectly catch the exception. Regards, Mike _______________________________________________ mkgmap-dev mailing list mkgmap-dev(a)lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

1 0

Re: [mkgmap-dev] Commit r4448: fix possible NPE introduced with r4432
by Mike Baggaley 15 Feb '20

15 Feb '20

>Version mkgmap-r4448 was committed by gerd on Thu, 13 Feb 2020 > >fix possible NPE introduced with r4432 >FileNotFoundException was ignored. May result in a NullPointerException when input file is not in the expected format. e.g. with >java -jar mkgmap.jar . Hi Gerd, I can't see how this update can make any difference. The change catches the FileNotFoundException and immediately re-throws it, so it effectively does nothing. Looking at the call hierarchy, FileNotFoundException is caught in createRawData, makeMap and loadCoastlines so all the functions that call the load function directly or indirectly catch the exception. Regards, Mike

2 1

Commit r4452: documentation improvements and program to generate help file options from doc/options.txt
by svn commit 15 Feb '20

15 Feb '20

Version mkgmap-r4452 was committed by gerd on Sat, 15 Feb 2020 documentation improvements and program to generate help file options from doc/options.txt Based on patches from Mike Baggaley To generate help/en/options from doc/options.txt use ant gen-options-file Take care: no backup is made, manual changes are overwritten http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4452

1 0

Fwd: Re: AW: More method options for is_in function
by Arndt Röhrig 15 Feb '20

15 Feb '20

2 3

Minor documentation update
by Mike Baggaley 14 Feb '20

14 Feb '20

Hi Gerd, now that my previous documentation patch has been committed, I have noticed a couple of errors in it that I didn't notice because I didn't have the means to compile it into the web pages. Please find attached a small patch that corrects these errors and adds some information on the input files, primarily to describe typ file compilation. I have also slightly tweaked the code for generating the options file from options.txt, but as the previous version does not appear to be in SVN, this has come across as a complete file, rather than the change (which is just the deletion of the hyphen character in 'if (line.startsWith(";-")) {'). Regards, Mike

2 3

More method options for is_in function
by Gerd Petermann 13 Feb '20

13 Feb '20

Hi all, see http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4442 @Ticker: I assume you are working on an alternative implementation of the methods in IsInUtil? If not I'd like to remove all the code duplication introduced with the last patch. Gerd

4 22

Commit r4450: Produce better error message when input-file is a directory
by svn commit 13 Feb '20

13 Feb '20

Version mkgmap-r4450 was committed by gerd on Thu, 13 Feb 2020 Produce better error message when input-file is a directory This also prevents a possible StringIndexOutOfBoundsException http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4450

1 0