Adding data curation tool to Harvard Dataverse

[stevenwinship]

Adding Data Curation Tool to https://dataverse.harvard.edu/ and https://demo.dataverse.org/

Closes #32

[pdurbin]

As I mentioned in Slack, there is Borealis-specific language in the tool (I check on our demo server):

[stevenwinship]

Removed the Data Curation tool from Demo. Putting this issue on hold until a new release is available

[stevenwinship]

Message From Nana Boateng:
We have a beta release planned for Tuesday August 27, 2024. We are setting up a test instance at demo.borealisdata.ca for testers.

[pdurbin]

@nana-boateng @bikramj any news on the release? I just checked https://github.com/scholarsportal/Dataverse-Data-Explorer/releases . We're just trying to figure out when it makes sense to pull this item into sprint. No rush. Thanks.

[nana-boateng]

@stevenwinship sincere apologies again for the late response, I just saw your email. Please reach me at my work email [email protected] ([email protected] is my personal email)

@pdurbin We are planning on releasing this at the end of November (the week of the 25th).

I have also updated the app to replace hardcoded references to Borealis, with 'Dataverse' this will be further customizable with language files on release.

[cmbz]

Placed on hold to check back in about the potential new release in November. Assigning to @stevenwinship and @cmbz for monitoring.

[cmbz]

2024/12/02: No release as of this date. Will check again on 2025/01/15.

[nana-boateng]

Hi all,

Sincere apologies for the delay. The new Data Explorer is now in pre-release and will be officially released later today.

[cmbz]

That's great new @nana-boateng, thanks for letting us know! :)

[cmbz]

2025/01/27: Moving back into Sprint Ready now that release 2.0 is available.

[stevenwinship]

@nana-boateng Could you let me know if this error is one that needs to be looked at or if it's ok to ignore. I see the tool making the call to dataverse and the data in the response looks good. I don't see anything other than a stripe of color on the screen. This file is definitely missing. I'm running in docker.

2025/07/17 18:10:45 [error] 36#36: *9 open() "/usr/share/nginx/html/matomo.js" failed (2: No such file or directory),

EDIT: got it working. Nothing to do with that error.

[stevenwinship]

@nana-boateng I have the tool running in demo.dataverse.org. I'm currently getting this error and I would like help fixing this.

Error fetching dataset: Bad signed URL

[stevenwinship]

This error is coming from dataverse. It is unable to get the user from the signed url. This works locally in docker. On demo it fails with user logged in with username/password or shib (Harvard Account).

[nana-boateng]

@stevenwinship looks like you figured it out. But let me know if you find anything else I can fix!

[qqmyers]

FWIW: Doing bin/asadmin set-log-levels edu.harvard.iq.dataverse.util.UrlSignerUtil=FINE (which is still on on demo), I see that Dataverse is trying to match the hash with String to hash: http://demo.dataverse.org/api/v1/files/2664681/metadata/2360434/toolparams/69?until=2025-07-18T16:29:47.724&user=qqmyers&method=GET&token=<key>]] - starting with http. Since demo is set up on https, this fails.

My guess, based on a problem with a TurboCurator demo around DCM2025, is that the Anubis configuration in /etc/httpd/conf.d/ssl.conf is still not sending the correct URL for Dataverse to pick up. I'm not sure yet how to fix that.

[stevenwinship]

@landreev Here is the issue

[stevenwinship]

Everything is working now. This tool was configured on qa.dataverse.org, demo.dataverse.org, and both nodes of https://dataverse.harvard.edu/