formatting user geometries

[HeatherSavoy-USDA]

Currently, the API expects the user geometries as strings in the following format:

Polygon endpoint:

Specifies the clip boundary for the subset operation. The boundary can be specified in one of two ways: 1) The upper left and lower right corners of a bounding box for subsetting the data, specifed as a comma-separated list of the form "(UPPER_LEFT_X,UPPER_LEFT_Y),(LOWER_RIGHT_X,LOWER_RIGHT_Y)."; 2) A comma-separated list of coordinates defining the vertices of a clip polygon as in "(X1,Y1),(X2,Y2)...".

Points endpoint:

The x and y coordinates of point locations for extracting from the data, specified as x1,y1;x2,y2;...

This package currently supports the first polygon format as a list of two coordinate pairs. This should be expanded to accept any of the following input, then extract+format coordinate info for the API (assuming endpoint from format):

• sf objects with POLYGON or MULTIPOINT geometry (polygon or point endpoint, respectively)
• sp SpatialPolygon* or SpatialPoint* objects (polygon or point endpoint, respectively)
• geojson polygon or multipoint objects (polygon or point endpoint, respectively)
• raster object (polygon endpoint)
• c(xmin,ymin,xmax,ymax) (polygon endpoint)
• data.frame or tibble with two columns (point endpoint)
• list of coordinate pairs (point endpoint)
• or any of the three string formats above

[HeatherSavoy-USDA]

API now accepts more geometry formats, including file upload.

[HeatherSavoy-USDA]

Current status: most but not all of these geometry types are supported.

Supported: sf and sp objects, geojson file, data.frame for both endpoints. + shapefiles and csv.

Not supported: vector or lists of coordinates, or strings to directly pass along to GeoCDL .