Fix harvesting script to get all points and boxes
Currently, the code that extracts spatial information from EML documents stored
in Metacat only grabs points and does not create polygons that better represent
the bounding boxes expressed in EML. Also, the code grabs the first of the
geographicCoverage sections, instead it should grab all.
Updated by Matthew Perry over 16 years ago
This bug is a few parts and is already partially resolved:
Bug 2499 describes the effort to port the harvesting script to pure Java. This has been relatively successful. It now generates two datasets representing each metacat document as a polygon AND a point. The point is calculated from the centroid of the geographic coverage.
Multiple geographicCoverage sections are not yet supported
Documents without geographicCoverages (or left blank) will are represented as a point with lat,long of 0,0