[Fwd: Re: WFS gurus . . . .]

[Bob Basques]

Frank,

This is exactly what we we're looking for.

I guess we just needed to dig far enough to get our terminology right to 
ask the question(s)

Thanks

bobb


Frank Warmerdam wrote:
> Bob Basques wrote:
>   
>> All,
>>
>> Some progress,
>>
>> We've managed to add to the PROJ4 EPSG list and verified that it works 
>> with a WMS output.  So the new EPSG code we added works just fine as a 
>> WMS call.
>>
>> Next we tried the same with a WFS call and got an INIT error, which was 
>> traced to OGR, so we then (tried to) build Mapserver without GDAL only 
>> to learn that WFS requires it to build.  So my question is, where is OGR 
>> grabbing it's Projection info from?  From the code it looks like there 
>> is a parser in place, but we couldn't find where it was being referenced.
>>     
>
> Bob,
>
> OGR creates it's EPSG coordinate system information from the pcs.csv and
> gcs.csv tables distributed with GDAL.  Note that periodically these
> tables are regenerated from the EPSG database, and these .csv files are
> the basis of generating the PROJ.4 epsg translation file.
>
> In theory you can extend the pcs.csv with your projections but the format
> is a bit esoteric.
>
> I believe you can add Well Known Text format definitions to the "epsg.wkt"
> file in gdal/data to add stuff not in the pcs.csv and gcs.csv files though
> I haven't used this mechanism for a while.  This is what I use to add
> Cubewerx and ESRI extensions to the EPSG dataset.
>
> See gdal/data/cubewerx_extra.wkt for examples of how the definitions would
> be formatted.
>
> Best regards,
>