November 2021 - mkgmap-dev - The mkgmap lists

Commit r4816: Use Normalizer.normalize() before converting unicode strings to label
by svn commit 16 Nov '21

16 Nov '21

Version mkgmap-r4816 was committed by gerd on Tue, 16 Nov 2021 Use Normalizer.normalize() before converting unicode strings to label Fixes Twülpstedt issue (u-umlaut either encoded as multiple characters u + "COMBINING DIAERESIS" (0x75 + 0x308) or as single character LATIN SMALL LETTER U WITH DIAERESIS (u+00fc) http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4816

1 0

Twülpstedt, Normalisation of unicode strings
by Gerd Petermann 16 Nov '21

16 Nov '21

Hi all, see also https://forum.openstreetmap.org/viewtopic.php?id=74231 mkgmap sometimes fails to encode correct strings for a given codepage like 1252 (latin1). I've uploaded a file that contains an area in Germany where the u-umlaut in name Twülpstedt is encoded in two different ways, either with ü (0xfc) or as u + "COMBINING DIAERESIS" (0x75 + 0x308) See umlaut.osm at https://files.mkgmap.org.uk/detail/537 With the current code the 2nd variant is displayed as Twu?lpstedt. This 1-liner name = Normalizer.normalize(name, Normalizer.Form.NFC); helps to change the name to the usual encoding which works well with the codepage translation. So far so good. Now I wonder where exactly this call should be placed. My first idea was the code where the string is converted to a Garmin label, but maybe it should happen much earlier so that also the style rules "see" the normalized form. Any thoughts? Gerd

3 6

change mdr25 logic
by Gerd Petermann 16 Nov '21

16 Nov '21

Hi Ticker, attached are two slightly different patches for Mdr25 which I think make the code more plausible. v2 would be my choice. I've not yet tried to find out (again) how the sort works regarding strength. In my maps I see that Mdr25 contains fewer records than Mdr5 (e.g. 7609 of 8275 or 1146 of 1850), but I have no clue yet how to find out which recrods from Mdr5 are skipped, means, what exactly has to be compared. Gerd

3 8

Commit r4815: remove unused import
by svn commit 15 Nov '21

15 Nov '21

Version mkgmap-r4815 was committed by gerd on Mon, 15 Nov 2021 remove unused import http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4815

1 0

Commit r4814: fix Mdr12 index position calculation for maps with mixed case labels
by svn commit 15 Nov '21

15 Nov '21

Version mkgmap-r4814 was committed by gerd on Mon, 15 Nov 2021 fix Mdr12 index position calculation for maps with mixed case labels Seems the position must refer to the first entry in Mdr11 that mmatches using PRIMARY strength. Alternative would be to stop writing this small section, but maybe it really helps on some devices http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4814

1 0

Problem with global index POI search
by Gerd Petermann 15 Nov '21

15 Nov '21

Hi devs, Enrico has contacted me in a PM about a rendering problem with his maps and while analysing that I found an obvious bug searching for e.g. "Scuola Dell Infanzia Fondazione Pietro Caprettini" as name of the POI without filling any further fiels in MapSource. When I start to type that name MapSource updates the list of candidates. For scuo it shows the full name "Scuola Dell Infanzia Fondazione Pietro Caprettini" but when I select that entry with the mouse the search returns "no items found" When I type scuol the name "Scuola Dell Infanzia Fondazione Pietro Caprettini" disappears from the candidates list as well as all other names starting with "Scuola dell". My understanding is that we probably create wrong Mdr11 entries, but MdrCheck doesn't complain. The tiles are at https://files.mkgmap.org.uk/detail/536 The command to reproduce the problem: java -jar mkgmap.jar --code-page=1252 --index --gmapi 66923003.img 66923005.img 66923006.img The tiles were created with the --lower-case option and a rather special style which sometimes adds leading / trailing blanks to names. It also produces lots of equally named POI. Gerd

3 7

Commit r4813: Format change to prevent error
by svn commit 08 Nov '21

08 Nov '21

Version mkgmap-r4813 was committed by steve on Mon, 08 Nov 2021 Format change to prevent error There was a line beginning with a '*' which is not meant to be a list and was causing a crash preventing the page being displayed on the website. Re-arranged the line so that it didn't start with the '*' http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4813

1 0

Commit r4812: fix possible but very unlikely error in mdr25 section
by svn commit 08 Nov '21

08 Nov '21

Version mkgmap-r4812 was committed by gerd on Mon, 08 Nov 2021 fix possible but very unlikely error in mdr25 section - handle the case that the same combination of city name and region name occurs in two or more different countries http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4812

1 0

dokumentation page reports error
by 7770 06 Nov '21

06 Nov '21

Hi. In case anyone reading this list is also the webmaster of mkgmap.org.uk. The page: https://www.mkgmap.org.uk/doc/options reports: A server error occurred We're sorry but there is currently a problem with the site. We apologize for any inconvenience, please try again later. Regards Karl

1 0

new splitter version r643
by Gerd Petermann 06 Nov '21

06 Nov '21

Hi all, see https://www.mkgmap.org.uk/websvn/revision.php?repname=splitter&rev=643 I've merged the branch now although the branch version sometimes produces slightly worse results compared to r642. As Felix pointed out one general problem with splitter is that the results for some countries or continents are not very stable. This also depends on the usage of polygons and the complexity of those polygons as well as the --precomp-sea option. I invested a lot of time this year to find better algorithms for the solver but without much success. So, I hope the changes don't cause any serious problems. If splitter fails to find a (good) split please report the used options and post a link to the file densities-out.txt. Gerd

2 1