search

nulib / arch

Northwestern University institutional repository, built on Samvera's Hyrax gem.
2 stars 0 forks source link

Add "Download All" to Multi-file Works #258

Closed chrisdaaz closed 6 years ago

chrisdaaz commented 7 years ago

Severity

Is the production site running?

Are staff blocked from performing their work?

Descriptive summary

Relates to Hyrax 1.x

Rationale

Users may want to download all of the files in a Work with one click, such as this example of MATLAB scripts and formatted data.

Story: Reader reads an article that lists the DOI of the article's supporting data set. Reader visits the data set in the institutional repository, represented as a Work. The Work contains X number of files. Reader wants to download a local copy of the data set.

Expected behavior

Here's an example from Figshare:

image

Here's an example from a journal:

image

Actual behavior

Download each file individually.

Related work

Created a Hyrax issue here: https://github.com/samvera/hyrax/issues/1203 Deep Blue Data example: https://deepblue.lib.umich.edu/data/concern/generic_works/3f462547h Deep Blue Data PR: https://github.com/mlibrary/umrdr/pulls?utf8=%E2%9C%93&q=is%3Apr%20is%3Aclosed%20download%20all

davidschober commented 7 years ago

@nulib/repodev it looks like the hyrax group is OK with an upstream PR. Does someone want to push Michigan?

chrisdaaz commented 6 years ago

bump. i've been asked about this by two major stakeholders, and i see that michigan has implemented this on a hyrax 1.0.5 app

davidschober commented 6 years ago

@csyversen @carrickr I'd like to investigate this as part of our current arch work. Can one of you take a look at this and give me an estimate? Can we implement Michigan's strategy?

davidschober commented 6 years ago

Chemical heritage foundation also executed this on a sufis-based app https://digital.sciencehistory.org/works/0g354f76p https://github.com/sciencehistory/chf-sufia/pull/1050

bmquinn commented 6 years ago

@chrisdaaz @kdid and I are working on this issue this sprint, and have a clarification question about which files to download. It looks like the SHI download only downloads images with public visibility in a work, and Deep Blue only the file_sets for a work (both non-recursively). Our first crack at it downloads all files in all file_sets nested within a work recursively (so it downloads the files in file_sets for a work and the files in file_sets for a child work of the first work). On second thought maybe that's too much? Not sure what the users' expectation would be for "download all". Should we be doing what Deep Blue is doing and just download the files of file_sets for the work and none for child works instead? Feel free to hit me up in the #sufia slack channel if you want to chat.