[OSM-dev] Osmosis Replication Statistics

Brett Henderson brett at bretth.com
Sat Sep 1 08:32:05 BST 2007


Martijn van Oosterhout wrote:
> By the way, look at the extract-polygon program in SVN.  It allows for
> multiple polygons as well as inclusions and exclusions. We use this a
> lot during the AND conversion process. If you want to use that polygon
> format it would make it easier...
>   
I can't read perl to save myself :-)  And I can't run that particular 
app without installing the math polygon library manually from CPAN so I 
couldn't get it to spit out the usage instructions.  But I did dig this 
section out of the program which describes the polygon format.

1
        -0.6450E+01     0.4980E+02
        -0.2000E+01     0.4890E+02
        -0.1850E+01     0.4925E+02
        -0.2080E+01     0.4973E+02
         0.1350E+01     0.5090E+02
         0.2250E+01     0.5258E+02
        -0.0500E+01     0.6130E+02
        -0.8920E+01     0.5785E+02
        -0.1140E+02     0.5130E+02
        -0.6450E+01     0.4980E+02
END

I assume that "1" and the "END" mark the beginning and end of sections.  
Does this mean that a subsequent polygon section would begin and end 
with "2" and "END" respectively?  And finally is it essential that all 
polygon files be broken into sections or do you allow a file with only a 
single polygon to omit the section header and footer?

I'm unsure how to get a java.text.DecimalFormat pattern to match the 
above format but I'll keep looking into it.






More information about the dev mailing list