Feature Request/Idea: custom metadata fields for files

[vkush]

Overview of the Feature Request For archives is very important to describe each file with maximum metadata as possible. Usually there are separate xml files with the metadata, which are stored together with the files, e.g. with such PREMIS blocks (and related subblocks) inside:

...
<premis:objectIdentifier>
<premis:fixity>
<premis:format>
  <premis:formatDesignation>
  <premis:formatRegistry>
  <premis:formatNote>
  ..
  <premis:formatNote>
...

It would be practical to have a possibility to create custom metadata fields also on the file level, similar as custom metadata blocks on the dataset level. Currently it is possible only to use the "File Name", "File Path" and "Description" fields - what covers only partially the available metadata in the archives.

What kind of user is the feature intended for? All users

What inspired the request? Requirement of archives to have metadata on the file level

What existing behavior do you want changed? Currently limited amount of metadata fields on the file level

Any brand new behavior do you want to add to Dataverse? Possibility to add/edit custom metadata for individual files - via GUI and API. Some metadata should be not edited, e.g. size, MD5 etc (as it realized currently).

Any related open or closed issues to this feature request?

• Same question in the community: https://groups.google.com/g/dataverse-community/c/Dbn3LdZyNn8/m/lglS6Do3BwAJ

• 2295

[pdurbin]

@vkush thanks for opening this issue! A few thoughts for you.

• In addition to file name, file path, and description, you can add tags (known in the code as categories), which could be considered a form of metadata: https://guides.dataverse.org/en/5.8/api/native-api.html#adding-file-metadata
• With our "sample data" repo we sometimes add what metadata we can to files. As we develop new ways to add metadata, it would be good to update examples such as https://github.com/IQSS/dataverse-sample-data/tree/acc675734d0fcd06b4a1d8862060777055742d6f/data/dataverses/open-source-at-harvard/datasets/open-source-at-harvard/.filemetadata . See also https://github.com/IQSS/dataverse-sample-data/issues/5
• For tabular data you could use/abuse the new and experimental "auxiliary files" feature to upload a file of your choosing along side each tabular file. It could be XML, JSON, plain text, etc. Please see https://guides.dataverse.org/en/5.8/developers/aux-file-support.html

[pdurbin]

@vkush did any of these suggestions help?

Also, here's another related issue:

• 6455