galterlibrary / inveniordm-at-nu Goto Github PK

It would be nice if we had something like a Docker Trusted Registry https://docs.docker.com/datacenter/dtr/2.2/guides/ where we could push our Docker images and have the same images be used from test to production.

The advantages:

• Environment consistency across CI and production
• Continuous delivery
• Faster builds because images can be reused

The disadvantages:

• Some complexity in setting this up
• Permissions / access control
• Service to maintain

As a researcher, I want to get search results from other cd2h repo instances when I search on my institution's cd2h repo instance.

[moved from #62]
As a researcher (or researcher proxy), I want to request access to indexed (but not necessarily uploaded) datasets from another researcher.

As a researcher, I want to receive an automatically generated email from the repository when another user submits a request for one of my datasets.

Zenodo has an access-request feature.

Probably copy from DataCAT?
Usage of https://schema.org/ models?

Related libraries:

• https://github.com/inveniosoftware/invenio-jsonschemas
• https://github.com/inveniosoftware/domapping (:warning: Experimental)

As a machine, I want to be able to rely on resource signposting to navigate and discover metadata and data related to a resource.

The following links in HTML head should be present:

Signposting resources:

• signposting.org

For the base invenio project and data-model

At least:

• MIT
• Creative Commons 4
• GPL
• whatever license given by institution

Ansible? Docker?

Deployment process must be documented and customizable because a cd2h-repo-project instance should be installable on a given institution's servers as simply as possible.

Ideally, deployment should be a one command action that will:

• Create database tables if not there
• Create (elastic)search indices if not there
• Run migrations that have not been run
• Update indices with potential changes
• Rebuild CSS/Javascript assets

The workflow to run one-off command should also be documented.

Part of #1

Part of #9

Related libraries:

• https://github.com/admiralobvious/flask-simpleldap
• https://github.com/inveniosoftware/invenio-userprofiles
• https://github.com/inveniosoftware/invenio-accounts
• https://github.com/inveniosoftware/invenio-access

As a researcher I want to find collaborators using similar research methods, domains, and subject pools in order to design collaborative projects.

*Note from 06/18/18 DevSci meeting: "All of DevSci's work is collaborative, spanning schools and units. Potential collaborators search for each other now in DevSci and have a hard time locating each other."

-Subject domains, research methodologies, tools, measurement techniques, demographics (sample-level metadata per CD2H 6/18/18 meeting) are all candidates for schema to be included.

Related libraries:

• https://github.com/inveniosoftware/invenio-orcid (:warning: experimental)
• https://github.com/inveniosoftware/domapping (:warning: experimental)
• https://github.com/inveniosoftware/invenio-userprofiles

Related libraries:

• https://github.com/inveniosoftware/invenio-userprofiles
• https://github.com/inveniosoftware/invenio-accounts
• https://github.com/inveniosoftware/invenio-access

I don't think we need to store users' NetID in db, but maybe we should? I can use the email from LDAP to find the user in the database. However we'd need to disallow setting custom emails in user profile (obvious security hole.) What do you think @fenekku?

Auto-verify user's email.

Part of #1

• Disable registration
• Disable change-password (link to NU password-change page instead)
• Disable the 'forgot-password' action
• Disable changing email, name and other attributes? (Depends on #43)
• Disable email verification (Handled in #43)

Part of #1

Related to #1

As a researcher, I want to be able to keep abreast of another researcher's activities.
As a researcher, I want to be able to stop being notified of a researcher's activities.

Events generated by the followed researcher should appear on the follower's activity feeds.

As an uploader/editor, I want to be able to not only add data to a given resource but also external links (e.g. a link to the code hosted on GitHub or a link to the spreadsheet on Google Drive)

Not a priority - even possible to contribute upstream

See https://docs.docker.com/compose/compose-file/compose-file-v2/#links

• As a user, I want to be able to upload a resource file. Done via #4

• As a user, I want to be able to upload or enter metadata without a resource file. Done via #7

• As a user, I want to be able to upload resource files with my metadata entries. Done via #4

• As a user, I want to provide and edit metadata for my uploaded files. (Aspect of this is an authorization issue). Editing of unpublished records done in #8 and #239.

• As a librarian, I want to provide and edit metadata for uploaded files. (This is more likely to be an authorization issue). Done via #250

• [ ] As a librarian, I want to be able to upload resource files to metadata entries.

As a researcher/record contributor, I'd like an automated way to solicit co-PI approval (through an email or other alert) before depositing datasets or records about them.

My LDAP library of choice is ldap3

As a researcher, I want to see how impactful or shared or viewed my research is.

In the user dashboard, have the same analytics as for a resource but combined across all resources of the user.

Related libraries: https://github.com/inveniosoftware/invenio-stats

The basic resource page should include:

• Title/Name of the resource
• Author
• Unique Identifier Moved to minting task
• Description

See https://digitalhub.northwestern.edu/files/a850ee99-3899-4381-a858-077c3b5ee03b for a more involved resource page

Related libraries:

• https://github.com/inveniosoftware/cookiecutter-invenio-datamodel

The form should contain:

• Agreement to terms of data upload before uploading

Part of #9

Related libraries:

• http://github.com/maxcountryman/flask-uploads/
• https://github.com/inveniosoftware/invenio-files-rest
• https://github.com/inveniosoftware/invenio-deposit (:warning: experimental)

There are a couple of improvements that can be made to have faster builds like:

• installing the dependencies before code change
• taking advantage of Docker's layer caching

Related libraries:

• https://github.com/inveniosoftware/invenio-iiif
• https://github.com/inveniosoftware/invenio-previewer
• https://github.com/inveniosoftware/invenio-deposit (:warning: experimental)
• https://github.com/silentsokolov/flask-thumbnails
• https://github.com/dysosmus/flask-thumbnails-wand

Separate out in 3 tasks.

Have file list contain a preview button to open up a modal with a preview of the corresponding file.

• #241
• #242
• #243

The goal is to easily configure various environments including development, testing and staging.

As a researcher, I want to be able to search records by the type of file they hold in order to pinpoint the results I am looking for faster.

This will confirm that file types are added to record metadata.

Assign a DOI to an entry on publication. We need to decide on our DOI format before we can do this:
See https://blog.datacite.org/cool-dois/ .

Related libraries:

• https://github.com/inveniosoftware/invenio-pidstore
• https://github.com/inveniosoftware/idutils
• https://github.com/inveniosoftware/datacite (I think it is related but maybe @phebal knows better)

As a depositor, I want my record to be assigned a unique and persistent URL/identifier, after I catalog so that my record is reference-able and long-lasting.

Versioning will be done later, so multiple publications do not change DOI.

• Read up on DOI
  • https://blog.datacite.org/cool-dois/
  • https://ezid.lib.purdue.edu/learn/id_concepts
  • https://support.datacite.org/docs/doi-basics
  • https://support.datacite.org/docs/api
  • https://support.datacite.org/docs/doi-fabrica
  • https://en.wikipedia.org/wiki/Digital_object_identifier
  • https://www.doi.org/factsheets.html
  • https://www.doi.org/hb.html
• Generate DOI (value)
• Register value + metadata with Datacite #269
• Show DOI URL according to best practices on the record page #277

Allow override of the default authentication through app.config

Part of #1

Check with Invenio if they're receptive to switching to pipenv.

• Install chromedriver or geckdriver, make it available to Jenkins and switch on E2E environment variable.
• Figure out missing libraries for a headless experience

We will install it via Github and keep versions via branches tags.

Very much related to #9

• Number times record is viewed
• Number times resources are downloaded (aggregate as per Invenio)
• Number citations -- altmetrics will take care of this for now, until a more robust approach can be used

Related libraries:

• https://github.com/inveniosoftware/invenio-stats

Related libraries:

• https://github.com/inveniosoftware/invenio-search-ui

Related libraries:

• https://github.com/inveniosoftware/invenio-files-rest
• https://github.com/inveniosoftware/cookiecutter-invenio-datamodel

Download all files in a zip in one go.

As a researcher/record contributor, I want citations automatically generated for my publications and/or datasets in MLA, APA, Chicago, Harvard, and Vancouver styles (as Google Scholar offers). I also want downloaders of my materials to be strongly reminded to cite them, either through pop-up reminders during download, a paragraph about citation in a use agreement, or by including a citation file with every download package. Moved to #73

Edit - Summary of discussions until Apr 10

Just APA, Chicago, MLA, Harvard and Vancouver for now.

Publisher should be: menRva. Galter Health Sciences Library & Learning Center.

• Generate APA
• Generate Chicago
• Generate Harvard
• Generate Vancouver
• Generate MLA
• Display UI

Invenio uses falsk-security (via invenio-accounts) to handle form-based (database) authentication and https://github.com/inveniosoftware/invenio-oauthclient for oath. We need to create a module that allows LDAP authentication and auto-registers users as needed. This will serve as the master ticket for this task.

As a researcher, I want to have my ORCID profile be tied to my activities/data in this software so that it is taken into account in my overall researcher profile.

As a researcher, I want to own my presence on this system and know that other researchers on the platform are who they claim to be.

Related links:

• https://invenio.readthedocs.io/en/latest/orcid-login.html
• https://members.orcid.org/api

Visually represent compliance to FAIR-TLC of uploaded data/metadata

See: https://zenodo.org/record/203295

Multiple user experiences are possible here:

As a researcher/uploader of resource, I want to have an expert to review my metadata input to maximize the discoverability of my research.

or be it opt-in:

As a researcher/uploader of resource, I want to choose to have an expert review my metadata input to maximize the discoverability of my research.

or completely offload the task

As a researcher/uploader of resource, I don't want to have to input a lot of metadata with my resource and would rather have a librarian add it after the fact, at the cost of not having my resource available right away.

As an institution user, I want to be able to have access to most features of the CD2H project using:

• Edge
• Firefox - and mobile version
• Chrome - and mobile version
• Safari - and mobile version

Internet Explorer 11- will be supported on a "as-needed" basis.