Bug #3541
closedclarify documentation on polygons and grings, indicate relationship to FGDC
0%
Description
The documentation in this section is sparse. The NEON project is currently evaluating EML, and has pointed out a discrepancy between FGDC and EML which should be either explained or the 2 specifications should agree.
FGDC: types that are equivalent to "datasetGPolygonOuterGRing" and "datasetGPolygonExclusionGRing" require a minimum of 4 points for polygons
EML: a) datasetGPolygonOuterGRing: requires 3 minimum, and b) datasetGPolygonExclusionGRing requires 1.
Reasoning: a) FGDC documentation states that a polygon needs to have 4 points, and the first and last should be identical. However, XMLSchema does not have a way to enforce this. So EML requires only 3, and the 4th is implied.
b) allowing the exclusionRing to have only one point means that a single point can be excluded, presumably, a single station.
EML's documentation for datasetGPolygonOuterGRing should make this clear. It also could recommend that xsl stylesheets that transform EML to FGDC should repeat the first lat/lon pair as the last when creating a list of points.
the alternative to both datasetGPolygonOuterGRing and datasetGPolygonExclusionGRing is gRingType, which is a string of ordered pairs. Presumably, this was supposed to be a field that could be directly translated to FGDC. If so, the documentation should state this, and recommended that a minimum of 4 pairs be included. However, if this field remains an unrestricted string, this is not enforceable with xmlschema.