Repository for the metadata created by using dmdScheme based metadata schemes.
- metadata
- DIRECTORY.yaml
- SCHEME1NAME_VERSION1
- metadata_as_xml_1_.xml
- metadata_as_xml_2_.xml
- metadata_as_xml_3_.xml
- ...
- SCHEME1NAME_VERSION2
- metadata_as_xml_1_.xml
- metadata_as_xml_2_.xml
- metadata_as_xml_3_.xml
- ...
- SCHEME2NAME_VERSION
- metadata_as_xml_1_.xml
- metadata_as_xml_2_.xml
- metadata_as_xml_3_.xml
- ...
- db.sqlite SCHEME1NAME_VERSION1.sqlite SCHEME1NAME_VERSION2.sqlite SCHEME2NAME_VERSION.sqlite
- db.sqlite SCHEME1NAME_VERSION1.sqlite SCHEME1NAME_VERSION2.sqlite SCHEME2NAME_VERSION.sqlite
As everything is text, a csv should be sufficient. Otherwise, an sqlite would be more robust, but not that suitable for github. For the moment, we should leave it as a csv for the moment and master.
This is a yaml file, with the following structure:
- top level:
NAME_VERSION
of the schemename
: Name of the schemeversion
: Version of the schemecreator
: creator of the schemeinfo
: Info to the metadata
The upload is not automated yet. A possibility would be to have an R package which is
- cloning the repo
- adding the metadata to the directory
- adding the metadata to the db.*
- add the relevant info tho the DIRECTORY.yaml file
- commits the changes
- pushes the changes to github