Support GigaDB as a data provider

[mih]

GigaDB (http://gigadb.org) is a large database of public domain data across a variety of topics. They use datacite for metadata (i.e. datacite can be queried for their metadata), each dataset has a DOI, there is no required authentication, data access is via plain FTP.

Demo dataset, randomly picked from fontpage:

http://gigadb.org/dataset/100397

FTP site:

ftp://penguin.genomics.cn/pub/10.5524/100001_101000/100397/

It looks as if all datasets are stored under this common URL:

ftp://penguin.genomics.cn/pub/10.5524/

The final part of a full URL comes from the respective DOI, not yet sure what the second to last part is. However, only four unique values exist for that.

[yarikoptic]

Note: has "partOf" mechanism for datasets collections: e.g. http://gigadb.org/dataset/100215 . Could be reflected as a hierarchy but not sure if should be