mik
mik copied to clipboard
Fedora3 toolchain
Create a toolchain that exports data from Fedora Commons 3 into ingest ready packages.
There may be situations where Fedora Commons 3 was used as the datastore for a digital repository ecosystem, but where the ability to use the metadata and file manipulators or other features of MIK, would help create better quality ingest packages than direct upgration.
If this toolchain is of interest to you, please give it a thumbs up so that I can gauge interest.
upgration++
A couple of ideas for approaching this toolchain.
First, if MIK has HTTP access to Fedora's REST API, the toolchain could ask Fedora to describe an object, then retrieve each of its datastreams, and write them to Islandora ingest packages for ingest.
Second, if MIK did not have access to the REST API, exporting the FOXML of the objects to be migrated would be an option. In this case, MIK would use the FOXML representation of each object to generate Islandora ingest packages. The "migrate" type of FOXML includes pointers to datastream content, whereas the "archive" type includes every part of an object in one FOXML file by encoding binary datastream content using base64 and wraping it in <foxml:binaryContent>
markup. The latter format may provide the best option if any of the datastream locations have access controls applied to them.
In either case, https://github.com/mjordan/islandora_batch_with_derivs may be an off-the-shelf batch module to use for the actual ingestion. All MIK would need to do is write out the data from Fedora in the directory structure documented on that module's README.