[mkgmap-dev] Commit r4271: improve reader for polish (*.mp) format
From Andrzej Popowski popej at poczta.onet.pl on Mon Feb 11 16:43:35 GMT 2019
Hi Gerd, I have complied a big map from mp sources (about 3.8GB in 50 files) in one go. Map is usable, routable, with index search and DEM. It is quite impressive, thanks! The problems I have noticed: - Overview map is not accepted by Mapsource and Basecamp. It looks ok, but it includes NET, NOD and DEM. Maybe empty NET/NOD is a problem? - I can't control parameters defining house number after street name or zip code before city name. I guess mkgmap takes them form some config file, which is not used for mp sources. - Value of mapid is taken form file name instead of mp source. - There seems to be 2 sets of city/region/country names in search index. In my sources country is named "POLSKA TOPO~[0x1d]PL" and some results from address search contain this name, while others use simply "PL". I don't use "bounds" option, I can't guess from where comes country "PL". In my mp sources, city/region/country appears in 2 places. Once as statements CitName, RegionName, CountryName, second as a part of Numbers# statement. Maybe there is different processing for them? - ~[0x1d] is a special character, indicating country/region abbreviation. See cgpsmapper manual paragraph 4.2.2.1. Mkgmap doesn't process this code in country name. My guess is that any value inside "~[]" could be simply inserted into label as a hex value. - For search results with country name from my sources, national characters are wrong. Maybe mkgmap assume that mp source is in Unicode? Cgpsmapper doesn't support Unicode, my mp sources are in code page 1250. -- Best regards, Andrzej
- Previous message: [mkgmap-dev] Commit r4271: improve reader for polish (*.mp) format
- Next message: [mkgmap-dev] Commit r4271: improve reader for polish (*.mp) format
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the mkgmap-dev mailing list