Replace current string approach with _resource objects. Affected fields include creator, contributor, publisher (maybe).

Do this in conjunction with reviewing current approach to _resource objects against new Hyrax functionality around controlled vocabs (geonames in particular) - see dog biscuits 39

Also see dog biscuits 38

This is no longer needed for KF.

CreateDerivatives is really slow. I'm also seeing LOADS of mini_magick* and magick* temp files in /tmp. Uploading 17 PDFs took ages.

See https://docs.google.com/document/d/1y9z5_USa2Y6WA5zexzJvWxyF2pGjgNygI-JACyTR-jg/edit for a log of issues. This is local deployment.

Things to test / re-test

• batch upload
• cloud providers
• contact form email
• setup groups and users, sharing works
• workflow design and load
• how to assign items automatically to different admin collections, by user_group or object model
• aws cloud template

• 6. Search Within (https://groups.google.com/forum/#!topic/universalviewer/wzCiQfpHn34)
• 7. Passing search into search within using this? in UV 3.0 (UniversalViewer/universalviewer#156)

Hyku uses iiif_manifest: https://github.com/samvera-labs/iiif_manifest

cf. Fedora Import/Export tool and pul Grocer gem.

When we have an 'index codes' txt file from eprints, this is a compressed version of all words in the extracted text, so it should be indexed in solr to complement and enhance anything we get from the full-text extraction in solr.

Cf eprints downloader code

This will give us a 'receipt' to verify the upload against.

csv with these: object id, database file id, filename

Current coversheet implementation draws the coversheet and combines with the pdf on the fly. This could be a performance hit under heavy load. Consider caching.

Would be generally useful. But directly useful for KF where private txt is first file added - this would be better moved to end.
Ability to resent thumb/representative in batch would also be good for KF where I want the first TIF to be rep.

Needs some tuning. Also see about running it as a service somehow. And more asynchronously?

Hyrax beta2 introduces a new notifications feature using Action Cable.

Sadly this can't be supported with Apache and Passenger: https://www.phusionpassenger.com/library/config/apache/action_cable_integration/

Will need to figure out how to deal with this, either switching to nginx or doing the reverse proxy do dah in the above.

Error is calling include? on NilClass. Easy fix.

Create a base script that can be reused across eprints repos with minimal work. Based on JSON export.

at the moment it's tied to required fields, but if this were a config then it would be easy to change without overriding code

(this would be a PR to Hyku)

• Model: include HasRendering and add rendering_ids to terms
• Form: def secondary_terms super - [:rendering_ids] end
• Presenter: < Hyku::ManifestEnabledWorkShowPresenter
• Controller: include Hyku::IIIFManifest

Adapt the eprints json importer to do MARC records

• 1. Only include public resources in public manifest (cf Hyku issue 1170)
• 2. Manifest should include (cf Hyku issue 1171):
  • 2.1 images only (phase one) - Hyku PR
  • 2.2 all UV-able resources (phase two)
  • 2.3 user configuration of manifest contents (phase three)
• 3. Expose more metadata in More Information - IN PROGRESS, see Hyku PR and iiif_manifest PR
• 4. Download whole PDF for an image-based object - IN PROGRESS, see Hyku issue 1372
• 5. Downloads not currently fully working (cf UV Issue 464) FIXED IN NEXT UV

Hyku uses iiif_manifest: https://github.com/samvera-labs/iiif_manifest

Turn HykuLeaf into a gem containing:

• base eprints importer
• some general purpose rake tasks
• generator for making a model into one that uses Universal Viewer

• 1. Only include public resources in public manifest (cf Hyku issue 1170)
• 2. Manifest should include (cf Hyku issue 1171):
  • 2.1 images only (phase one) - Hyku PR
  • 2.2 all UV-able resources (phase two)
  • 2.3 user configuration of manifest contents (phase three)

Hyku uses iiif_manifest: https://github.com/samvera-labs/iiif_manifest

Move / refactor the one in: https://github.com/ULCC/dart_hyrax/blob/master/lib/tasks/ulcc.rake

Need to run migrations as part of setup

Current code catches error when method doesn't exist. Would be cleaner to use receives to test if method exists, then log a WARN.

LoadError:
 cannot load such file -- importer

if running rspec tests individually ... must be something in spec_helper?

Current Hyku AWS demo is only running in dev mode.

Would be nice (possibly necessary!) to be able to swap out RIIF and use IIP. Look at how to do this and see if it would be easy to add as a configuration option.

Q. in terms of IIP pointing to files, is there any reason why IIP couldn't just point at fcrepo.object.directory ? after all, it's easy to find the file using the UUID (and via FileSet from solr using the digest)

Hyku uses RIIIF and so makes certain assumptions, eg. the path to images (/images/fileset/files/file) and the hostname for the IIIF image service (same as the app).

Files to look at in Hyku:

• https://github.com/samvera-labs/hyku/blob/master/app/presenters/concerns/displays_image.rb
• https://github.com/samvera-labs/hyku/blob/master/app/services/concerns/iiif_thumbnail_paths.rb
• https://github.com/samvera-labs/hyku/blob/master/app/presenters/hyku/manifest_enabled_work_show_presenter.rb
• https://github.com/samvera-labs/hyku/blob/master/app/builders/manifest_helper.rb
• https://github.com/samvera-labs/hyku/blob/master/config/initializers/riiif.rb
• https://github.com/samvera-labs/hyku/blob/master/config/routes.rb

At the very least list out the generators and tasks in here.

Single-value elements are adding 'empty' values to Fcrepo, and then showing up in the display interface.

Two solutions:

1. Fix it, eg. add something to the actor to delete any attributes that equal "" (or " ") before save.
2. Use multi-valued for everything

When updating the extracted text, I'm seeing the following error. This happens when the original file is TXT. I have observed it when adding to a PDF original_file, but this works more consistently than not.

No problem when I do a direct add of file into 'files' container via GUI and via CURL (as expected)

This is the code:

local_file = Hydra::Derivatives::IoDecorator.new(File.open(path, "rb"))
          local_file.original_name = path.split('/').last
          local_file.mime_type = Hydra::Works::DetermineMimeType.call(local_file, local_file.original_name)
          Hydra::Works::AddFileToFileSet.call(fileset,
                                              local_file,
                                              type.to_sym,
                                              versioning: false)

It works via this very manual method:

• fileset.association(:extracted_text).build
• PUT the file to the generated uri with curl using the FCR REST API
• add the ExtractedText rdf:type via api or gui

I've ruled out the being the file itself and also that it's a Fedora problem proper.

I can add this file to one fileset, but not the other, which hints at something to do with the original_file file in the fileset, which is a plain text file. It seems like it's something the AF / LDP are trying to do but I'm struggling to figure out is WHAT AF / LDP are doing ... I get nothing in the logs for any of this.

In FCRepo logs:

INFO 15:06:37.142 (FedoraLdp) PUT resource 'dev/6d/9c/fd/d2/6d9cfdd2-a775-4eb7-9575-f53f816f33f0/files/24eee720-d0a4-4ab3-a345-4d32bcdba79d'
DEBUG 15:06:37.142 (FedoraBinaryImpl) Created content node at path: /dev/6d/9c/fd/d2/6d9cfdd2-a775-4eb7-9575-f53f816f33f0/files/24eee720-d0a4-4ab3-a345-4d32bcdba79d/jcr:content
ERROR 15:06:57.163 (RepositoryExceptionMapper) Caught a repository exception: java.net.SocketTimeoutException: Read timed out

In Hyku logs:

Ldp::HttpError: STATUS: 500 org.modeshape.jcr.value.binary.BinaryStoreException: java.net.SocketTimeoutException: Read timed out
	at org.modeshape.jcr.value.binary.FileSystemBinaryStore.storeValue(FileSystemBinaryStore.java:128)
	at org.modeshape.jcr.value.binary.AbstractBinaryStore.storeValue(AbstractBinaryStore.java:251)
	at org.modeshape.jcr.value.binary.BinaryStoreValueFactory.create(BinaryStoreValueFactory.java:257)
	at org.modeshape.jcr.value.binary.BinaryStoreValueFactory.create(BinaryStoreValueFactory.java:49)
	at org.modeshape.jcr.JcrValueFactory.createBinary(JcrValueFactory.java:149)
	at org.modeshape.jcr.JcrValueFactory.createBinary(JcrValueFactory.java:41)
	at org.fcrepo.kernel.modeshape.FedoraBinaryImpl.setContent(FedoraBinaryImpl.java:178)
	at org.fcrepo.http.api.ContentExposingResource.replaceResourceBinaryWithStream(ContentExposingResource.java:612)
	at org.fcrepo.http.api.FedoraLdp.createOrReplaceObjectRdf(FedoraLdp.java:361)

See samvera-tech post

The invitation link will lowercase the email in the background, so running 'make me admin' will fail to find the user if the email has capitals in it. So ... lowercase the input.

Check the add users script and make sure those get invited as lowercase (I suspect they will).

Currently I have a conflict in the new migration added for shib with a duplicate column. This is caused (I believe) by the new invitible Hyku code, and should be easily resolved.

Also need to test out invitible and see if it plays with shib work.

• oai_dc support with sensible defaults which can be configured
• sets

See existing work:
https://github.com/code4lib/ruby-oai
https://github.com/projectblacklight/blacklight_oai_provider
https://github.com/osulibraries/blacklight_oai_provider (more up to date than ^ )
https://github.com/CottageLabs/blacklight_oai_provider
https://github.com/IISH/oai4solr

Change the invitible generator so that it removes 'registerable' and adds in a commented line to that effect.

Create an importer to update existing works with files.

Investigate whether this is happening already via the 'indexcodes' txt files.

If not, make it work.