[Imports] TMC LCL - automatic script

marcus.wolschon at googlemail.com marcus.wolschon at googlemail.com
Tue Nov 10 08:46:40 GMT 2009


On Mon, 9 Nov 2009 23:41:01 -0800, Sam Vekemans
<acrosscanadatrails at gmail.com> wrote:
> On Mon, Nov 9, 2009 at 11:02 PM, <marcus.wolschon at googlemail.com> wrote:
> 
>> On Mon, 9 Nov 2009 21:43:17 -0800, Sam Vekemans
>> <acrosscanadatrails at gmail.com> wrote:
>> > To understand,
>> > is that bounding box/ polygon something that came from the source?
>> > Or is it just the extents of the data?
>>
>> It is the bounding-box of all elements of the TMC LocationCodeList
>> that state this TMC-area as their administrative area.
>> It needs to be completely inside 2.6x the area of the OSM-Multipolygon/
>> Poylgon so we know to not have a city of the same name in a wrong
country
>> or state.
>>

> Isn't that the purpose of having the 'is_in' tag?

1: Only very few objects have this tag. So I cannot require it.
2: In the same state there can be 2 villages with the same name.
3: I'm trying to match that I have the correct area. That it covers about
   the same geographic region is a sensible test to do.

So, please no discussion about is_in here. This has nothing to do with
the sanity-check.

Any constructive comments about different sanity-checks and matching-rules
I could add/change to find more of the areas automatically?
Currently from the test-runs I estimate that in about 1/10th of the cases
it can find a match.

Marcus




More information about the Imports mailing list