Hi Carlos,
I tried your input file with the default style + your extra lines. I think I can reproduce
the problem, but my device shows two entries for the cities "Caceres" and "Cáceres".
When I select the first one I find almost no road, when I select the 2nd I seem to find every
road. I looked at the input file and found e.g.
http://www.openstreetmap.org/api/0.6/way/246419849
with
<tag k="addr:city" v="CACERES"/>
and
<tag k="addr:street" v="Calle Sierpes"/>
When I search for the road "Calle Sierpes", I find it in both cities,
for Caceres I find the entry produced by the way above,
for Cáceres I find other entries.
I tried it with --latin1 and with --unicode option, both work.
I think you get the correct result, it is just not what you want
because the data is wrong. Do you agreee?
BTW: For the given example and with my bounds file from 2015-11-13
I see mkgmap:admin_level8=Cáceres
so you would not have the problem if you move your rules after the standard
rule.
Gerd
Hi Carlos,
I noticed this problem before in other situations. My understanding
is that city names are collected from different sources, sometimes we
have additional info from a POI, sometimes we don't. If I got that right
the index will have two entries when a city name is derived from e.g. the
housenumber processing. I don't know why this is done, but their is special
code for this, so it seems to be intended.
I don't know why you see the problem only with the extra rules. Maybe
a different spelling in e.g. one addr:city ?
Gerd