Hi all,
I've been following some of this discussion about harvesting metadata, etc.
At Unidata we have been working on a new project called RAMADDA:
http://www.unidata.ucar.edu/software/ramadda
RAMADDA is a content management/publishing platform with a focus on
earth science data. You can think of it as "youtube for data". Our
approach is to combine content managment, publishing, collaboration and
data services all under one framework.
RAMADDA provides a number of harvesting and publishing interfaces, has a
rich underlying database backed content model and supports a number of
views of the content (html, thredds catalogs, rss, etc.) RAMADDA
incorporates many of the TDS services (opendap, grid subsetting,
catalogs) but provides a wide range of other services.
RAMADDA manages its content in a database but can also provide dynamic
views of a file system (like the TDS catalog generation) :
http://motherlode.ucar.edu/repository/entry/show?entryid=a136862d-1253-4096-b864-ac53028253d7
and can also provides these types of views on remote FTP sites. e.g. :
http://motherlode.ucar.edu/repository/entry/show?entryid=4d396a2a-7b09-4e3b-b3ad-165af41026b7
Furthermore, RAMADDA is a framework that allows one to easily extend the
core type system, add new kinds of metadata and add new views of this
content and metadata.
One of the facilities RAMADDA provides is the ability to harvest
metadata from a data file using the NetCDF/CDM stack. The time range,
spatial bounds, variables and global attributes are all harvested. For
example here is a radar file that has been harvested:
http://motherlode.ucar.edu/repository/entry/show?entryid=30f6f39c-3f38-4600-8d2e-70500f97f707
We recognize the need to be able to make use of the harvested metadata
from a file entry (i.e., a "granule") and copy it into a parent data
set element. So, we provide a web interface when editing an entry's
metadata and allow the user to simply copy selected metadata elements to
the parent dataset.
-Jeff
