[MetaCRS] Specifying Coordinate Order in Test Data File...Again

[Norm Olsen]

Hello Martin . . .

>> I suppose it's really an issue for the library itself to try and figure out this information somehow

Yes, that's exactly the point.  It is the library that maintains and interprets the definition.  Many use EPSG, but not all. Thus, if the "srcAuth" field is PROJ4, then the "srcCRS" references a PROJ4 definition of the coordinate system.  Thus, whatever the PROJ4 coordinate system definitions says, or doesn't say, about the axes is what controls for that particular test case.

If the source authority were EPSG, than the ordinates would need to be ordered in the manner that the EPSG definition specifies.  Yes, a library which does not honor axes swapping will fail if it tries to do certain EPSG systems which require swapping.  In my mind, that's part of what's being tested. 

Norm


-----Original Message-----
From: metacrs-bounces at lists.osgeo.org [mailto:metacrs-bounces at lists.osgeo.org] On Behalf Of Martin Davis
Sent: Monday, April 26, 2010 6:41 PM
Cc: metacrs at lists.osgeo.org
Subject: Re: [MetaCRS] Specifying Coordinate Order in Test Data File...Again

Makes sense.  So in this case, since the projection is ESPG:2227, the
ordinate order should be XY.

Am I right in thinking that some CS specification formats don't encode
the ordinate explicitly (eg ESRI, PROJ4) ?   Not sure if that's an
issue for the test files or not - I suppose it's really an issue for
the library itself to try and figure out this information somehow (eg
by assumption or exception)

Martin

On Mon, Apr 26, 2010 at 4:56 PM, Norm Olsen <norm.olsen at autodesk.com> wrote:
> Hello All . . .
>
> As I recall, the order of the ordinates in the test suite CSV file were to be the same as the system definitions being used as part of the test definition.  If the source CS has swapped axes, then the source ordinates would be swapped.  Similarly with the target ordinates.  Thus, the test suite can also be used to test axis ordering without concern about ordinate ordering.
>
> If you have access to something like Excel, swapping entire columns in the .CSV file is a simple cut and paste.
>
> Norm
>
_______________________________________________
MetaCRS mailing list
MetaCRS at lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/metacrs