Update GSD custom fields #2310
Comments
|
Committed changes, passing to Phil |
|
Give back to Sonia to check to check in dvn-build |
|
@sbarbosadataverse as we discussed I did my usual check to make sure the tsv change didn't require a change to the Solr schema.xml and it didn't. We sat and previewed the change on my laptop but we're going to wait to merge the 2310-GSD branch (currently just commit df60357 ) into 4.0.2 until you've heard back from Janina that the information you copied from the UI reflects the intended change. Passing this back to you until you've heard back. |
|
@sbarbosadataverse after you've confirmed with Janina that the change is ok, please pass this issue to @sekmiller who will write up instructions for @kcondon for what to do with the updated tsv file. |
|
@sbarbosadataverse ok, please feel free to update the branch we started: https://github.com/IQSS/dataverse/tree/2310-GSD |
for both class number and faculty list
|
@scolapasta |
|
@sbarbosadataverse just a heads up that we'll need to do the same thing as last time. I'll pull in your latest change to the GSD metadata block and I'll have you look at it to see if it's what you want before we merge it into 4.2. Let's coordinate a time to do this. |
|
Sounds good. Thanks
|
|
On https://shibtest.dataverse.org loaded customGSD.tsv from 4.1. Then I try to re-load the version from https://github.com/IQSS/dataverse/blob/76496aa9593736d7846e1aa0222fe229198762d5/scripts/api/data/metadatablocks/customGSD.tsv but I got this error: [root@dvn-vm3 api]# curl http://localhost:8080/api/admin/datasetfield/load -X POST --data-binary @data/metadatablocks/customGSD.tsv -H "Content-type: text/tab-separated-values" Here's the stack trace (I build the 4.2 war file on my laptop from commit 9ae1a64): Line 370 is As I mentioned to @scolapasta yesterday, the change in 76496aa (the version of the tsv I'm trying to load) seems to affect basically the entire tsv file. It's a much bigger change than the earlier commit at df60357 which seem to only change controlled vocabulary values. In short, I think there's something wrong with the latest version of the tsv file. The code that parses this tsv file is picky and I don't know much about it. @scolapasta was the original author and @sekmiller added the feature to re-load an updated tsv file (I'm not sure which issue that was). @posixeleni knows a lot about these tsv files as well. Again, I'm pretty sure we need a new one. One that doesn't cause the code to throw exceptions. |
I spoke with @sekmiller and he indicated that the concept of an identifier for a controlled vocabulary value was added after 4.0. Judging from #947 it was added in 4.0.1 by @scolapasta and @posixeleni . We were speculating that perhaps the problem is that the "identifier" column was empty but after uploading the version from 4.1 to Google Docs, I don't think that's the problem because the "identifier" column was empty back in 4.1 too:
That screenshot comes from here (4.1 version of the GSD block): https://docs.google.com/spreadsheets/d/1xQ8wi1-2NqylgzROf72A64ojrpQJAJTHdFe3mRzPHN0/edit?usp=sharing In addition the "journals" metadata block in 4.1 didn't have "identifier" filled in:
So I'm pretty sure "identifier" is optional. |
|
I spoke with @scolapasta and he indicated I should give this issue to @posixeleni to review the tsv file that is failing to import. @posixeleni here's the file: https://github.com/IQSS/dataverse/blob/76496aa9593736d7846e1aa0222fe229198762d5/scripts/api/data/metadatablocks/customGSD.tsv If there's something obvious you can fix, please feel free to push to the branch we've been using: https://github.com/IQSS/dataverse/commits/2310-GSD If it helps, I think the line with "01321" may be a problem. It's different than the surrounding lines: This is just a theory though... |
|
@pdurbin @sbarbosadataverse there's a bigger issue here than just the error coming out of this tsv. I can clean that up easily but it appears that the GSD wants us to replace their Course Names with new course names which after speaking with @sekmiller our code is not currently able to do yet so we at the moment we can only add new values to the tsv's controlled vocabulary. |
|
Waiting to get an ETA from @sekmiller on when we can replace controlled vocabulary values and not just add new values. |
|
@sekmiller as we discussed, you're welcome to look at adding a "preview" mode while you're in that part of the code: #2551 |
pdurbin
added
User Role: Curator
|
I'm going to close this very old issue as I think it pertains to a custom metadata block in dataverse.harvard.edu. I contemplated bringing this into a larger metadata consolidation issue (#6030) but I'm not sure what it's about (versioning metadata blocks, maybe?). |
|
I think this started as an update to a custom metadata block on Harvard Dataverse where the people using the metadata block wanted to edit the terms in one of its controlled vocabularies (instead of or in addition to adding terms). Maybe because the names of faculty changed or had typos. But the process of updating the metadata blocks doesn't handle editing terms in the controlled vocabularies. So I'm guessing that when this edited GSD metadata block was being uploaded, Dataverse saw the new controlled vocabulary terms in the tsv file and said, "Hey! There are saved datasets that have terms in the Faculty Name field that aren't in this new tsv file. You can't do that." Is that right? If so:
|
Janina changed custom field names
The text was updated successfully, but these errors were encountered: