Groups keyboard shortcuts have been updated
Dismiss
See shortcuts

custom metadata blocks - uploading & registering??

309 views
Skip to first unread message

Janet McDougall - Australian Data Archive

unread,
Sep 29, 2017, 1:57:20 AM9/29/17
to Dataverse Users Community
hi All
We are building some custom metadata blocks and vocabularies - I have been using the doco written up by Tim DiLauro, Johns Hopkins University - thanks.

I will test next week but I've seen a reference to "Register TSV metadata block" on https://github.com/IQSS/dataverse/issues/4037.   This is the first reference I've seen registering a TSV metadata block. 

Can anyone please add any further instructions for uploading and using custom metadata blocks in the datafile metadata beyond the very helpful doco by Tim?

thanks
Janet

julian...@g.harvard.edu

unread,
Sep 29, 2017, 6:30:38 AM9/29/17
to Dataverse Users Community
Hi Janet,

It would be perfect if the instructions you need (and hopefully get soon!) are shared so that I could help add them to that document Tim wrote up. I pinged someone from GESIS who may be able to help.

Thanks,
Julian

Philip Durbin

unread,
Sep 29, 2017, 6:57:56 AM9/29/17
to dataverse...@googlegroups.com
Hi Janet,

I believe "Register TSV metadata block" means to call the "/api/admin/datasetfield/load" API endpoint, passing the TSV file you created. For example, here's a link in the reference script Harvard Dataverse uses to load the custom metadata block for the Murray Research Archive:

curl http://localhost:8080/api/admin/datasetfield/load -X POST --data-binary @data/metadatablocks/customMRA.tsv -H "Content-type: text/tab-separated-values"

That's from https://github.com/IQSS/dataverse/blob/v4.8/scripts/api/setup-optional-harvard.sh#L47

For what it's worth, a while ago, I was wondering to myself it's a good idea to have a "preview" mode for loading a custom metadata block because they aren't easily deleted. I had opened https://github.com/IQSS/dataverse/issues/2551 about this.

I hope this helps!

Phil

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/57b59e45-f2dd-4832-b881-f6ad0dacb881%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--

Sherry Lake

unread,
Sep 29, 2017, 11:25:52 AM9/29/17
to Dataverse Users Community
I haven't added a new metadata block, yet (am working on one to support cultural objects as LibraData will be getting more and more 3D data files), but I have tweaked the citation metadata block.

YES, I know must be very careful doing that!! My tweaks have only been to change the description (cursor hover text) and watermark text.

I've used the curl command Phil has below, to register my changes.

Phil, I wold LOVE to have a "preview" mode for metadata before loading into the database.

--
Sherry


On Friday, September 29, 2017 at 6:57:56 AM UTC-4, Philip Durbin wrote:
Hi Janet,

I believe "Register TSV metadata block" means to call the "/api/admin/datasetfield/load" API endpoint, passing the TSV file you created. For example, here's a link in the reference script Harvard Dataverse uses to load the custom metadata block for the Murray Research Archive:

curl http://localhost:8080/api/admin/datasetfield/load -X POST --data-binary @data/metadatablocks/customMRA.tsv -H "Content-type: text/tab-separated-values"

That's from https://github.com/IQSS/dataverse/blob/v4.8/scripts/api/setup-optional-harvard.sh#L47

For what it's worth, a while ago, I was wondering to myself it's a good idea to have a "preview" mode for loading a custom metadata block because they aren't easily deleted. I had opened https://github.com/IQSS/dataverse/issues/2551 about this.

I hope this helps!

Phil
On Fri, Sep 29, 2017 at 6:30 AM, <julian...@g.harvard.edu> wrote:
Hi Janet,

It would be perfect if the instructions you need (and hopefully get soon!) are shared so that I could help add them to that document Tim wrote up. I pinged someone from GESIS who may be able to help.

Thanks,
Julian

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Philip Durbin

unread,
Sep 29, 2017, 12:28:12 PM9/29/17
to dataverse...@googlegroups.com
I'm glad you like the idea, Sherry. Can you please create a GitHub issue about adding a preview mode?

On Fri, Sep 29, 2017 at 11:25 AM, Sherry Lake <shla...@gmail.com> wrote:
I haven't added a new metadata block, yet (am working on one to support cultural objects as LibraData will be getting more and more 3D data files), but I have tweaked the citation metadata block.

YES, I know must be very careful doing that!! My tweaks have only been to change the description (cursor hover text) and watermark text.

I've used the curl command Phil has below, to register my changes.

Phil, I wold LOVE to have a "preview" mode for metadata before loading into the database.

--
Sherry

On Friday, September 29, 2017 at 6:57:56 AM UTC-4, Philip Durbin wrote:
Hi Janet,

I believe "Register TSV metadata block" means to call the "/api/admin/datasetfield/load" API endpoint, passing the TSV file you created. For example, here's a link in the reference script Harvard Dataverse uses to load the custom metadata block for the Murray Research Archive:

curl http://localhost:8080/api/admin/datasetfield/load -X POST --data-binary @data/metadatablocks/customMRA.tsv -H "Content-type: text/tab-separated-values"

That's from https://github.com/IQSS/dataverse/blob/v4.8/scripts/api/setup-optional-harvard.sh#L47

For what it's worth, a while ago, I was wondering to myself it's a good idea to have a "preview" mode for loading a custom metadata block because they aren't easily deleted. I had opened https://github.com/IQSS/dataverse/issues/2551 about this.

I hope this helps!

Phil
On Fri, Sep 29, 2017 at 6:30 AM, <julian...@g.harvard.edu> wrote:
Hi Janet,

It would be perfect if the instructions you need (and hopefully get soon!) are shared so that I could help add them to that document Tim wrote up. I pinged someone from GESIS who may be able to help.

Thanks,
Julian

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/283493d6-fcb0-4c7b-97f2-d7f3c1b8b034%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Janet McDougall - Australian Data Archive

unread,
Oct 2, 2017, 9:21:00 PM10/2/17
to Dataverse Users Community
ok re 'register'.  I just wasn't sure if it was an extra step beyond the CURL cmd.  Will share our outcomes, thanks all.
J


On Friday, 29 September 2017 20:57:56 UTC+10, Philip Durbin wrote:
Hi Janet,

I believe "Register TSV metadata block" means to call the "/api/admin/datasetfield/load" API endpoint, passing the TSV file you created. For example, here's a link in the reference script Harvard Dataverse uses to load the custom metadata block for the Murray Research Archive:

curl http://localhost:8080/api/admin/datasetfield/load -X POST --data-binary @data/metadatablocks/customMRA.tsv -H "Content-type: text/tab-separated-values"

That's from https://github.com/IQSS/dataverse/blob/v4.8/scripts/api/setup-optional-harvard.sh#L47

For what it's worth, a while ago, I was wondering to myself it's a good idea to have a "preview" mode for loading a custom metadata block because they aren't easily deleted. I had opened https://github.com/IQSS/dataverse/issues/2551 about this.

I hope this helps!

Phil
On Fri, Sep 29, 2017 at 6:30 AM, <julian...@g.harvard.edu> wrote:
Hi Janet,

It would be perfect if the instructions you need (and hopefully get soon!) are shared so that I could help add them to that document Tim wrote up. I pinged someone from GESIS who may be able to help.

Thanks,
Julian

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Janet McDougall - Australian Data Archive

unread,
Oct 3, 2017, 1:59:09 AM10/3/17
to Dataverse Users Community
hi All
My TSV custom metadata block registered "OK".  The only extra info I needed was to add it to the root dataverse (General Information->Metadata Fields) I had loaded it to.
Once i had done this it became available.  

I was lucky that my syntax worked, but I agree it would be helpful to validate the syntax or have a preview before applying to the database.  Do we need to create entire new Metadata Blocks if we want to change an existing one, OR are we able to make changes to existing Metadata blocks by loading (registering) again?  I can test tomorrow but be good to hear experiences.
 
Thanks for all the help.
Janet

Philip Durbin

unread,
Oct 3, 2017, 2:42:52 AM10/3/17
to dataverse...@googlegroups.com
In some cases, a metadata block can be reloaded. For example, the 4.8 release notes give instructions the following instructions:

"Update citation.tsv to add grant agency and number to facet and advanced search."

https://github.com/IQSS/dataverse/releases/tag/v4.8

I'd like to echo's Janet's call for others' experiences in this area.

Phil

Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/6e00e6d7-bc75-465a-9d55-0951d7c05ba7%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Janet McDougall - Australian Data Archive

unread,
Oct 3, 2017, 7:33:30 PM10/3/17
to Dataverse Users Community
Hi Phil
I have a question about the purpose of displayFormat: special variables. There is a comment in Tim's doco by Bill McKinney "Another potential point of confusion. This only will be used when the particular field is a child of a parent (compound metadata)".  Can anyone provide some more info about this? 

I will create some keyword vocabs  in stand alone Custom Metadata Block and see how it goes, will post outcomes.

Thanks
Janet  
Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Philip Durbin

unread,
Oct 3, 2017, 8:31:55 PM10/3/17
to dataverse...@googlegroups.com
Maybe four examples of "displayFormat" values from the datasetfieldtype* table will help:

 #NAME: #VALUE

 <img src="#VALUE" alt="#NAME" class="metadata-logo"/><br/>

 <a href="#VALUE">#VALUE</a>

 #VALUE

Hopefully, these examples demonstrate that there is some flexibility in how metadata fields are displayed. Sometimes you want to show more than just the value.

Is this helping? Obviously, this needs to be documented, which is what https://github.com/IQSS/dataverse/issues/3168 is about.

Phil

Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/46664ef8-4af8-4de1-929d-338f38723249%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Durand, Gustavo

unread,
Oct 3, 2017, 8:52:19 PM10/3/17
to dataverse...@googlegroups.com
Also, recent changes were made to use something like this for non compound. I can say more about this next week (at a conference* this week at presenting in a hour!)

* Phil is here too, but he's better than me at paying attention to e-mails. :)


On Tue, Oct 3, 2017 at 8:31 PM, Durbin, Philip <philip...@harvard.edu> wrote:
Maybe four examples of "displayFormat" values from the datasetfieldtype* table will help:

 #NAME: #VALUE

 <img src="#VALUE" alt="#NAME" class="metadata-logo"/><br/>

 <a href="#VALUE">#VALUE</a>

 #VALUE

Hopefully, these examples demonstrate that there is some flexibility in how metadata fields are displayed. Sometimes you want to show more than just the value.

Is this helping? Obviously, this needs to be documented, which is what https://github.com/IQSS/dataverse/issues/3168 is about.

Phil
Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/46664ef8-4af8-4de1-929d-338f38723249%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

Pete Meyer

unread,
Oct 4, 2017, 10:38:30 AM10/4/17
to Dataverse Users Community
A little more about the recent changes for non-compound fields.  One example we have in a custom metadata block: name="pdb", fieldType="text", displayFormat="<a href='http://www.rcsb.org/pdb/explore.do?structureId=#VALUE'>#VALUE</a>" (for reference, this is an identifier to a related entry in another repository).  A user would enter a valid identifier in the field (ex "3FKI"), and the display logic shows this as a html link "<a href='http://www.rcsb.org/pdb/explore.do?structureId=3FKI'>3FKI</a>" (aka - link text is "3FKI", but it's clickable).  Hopefully that doesn't confuse things more.

Best,
Pete
Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Janet McDougall - Australian Data Archive

unread,
Oct 9, 2017, 10:47:12 PM10/9/17
to Dataverse Users Community
Hi Phil and All

I have loaded several custom metadata blocks on a dev server to test how we want to collect certain metadata.  The outcome is I have a better idea of how to collect the metadata we need,  so I want to delete all the custom blocks and start from a clean slate.  We have now come across the difficulty of deleting loaded and reloaded custom blocks. 

Can anyone provide information on how to delete all references to custom metadata blocks from this Dataverse so there is no conflict with fields?   The only reference I've seen is that it's 'difficult'...

Thanks a lot
Janet

On Tuesday, 3 October 2017 17:42:52 UTC+11, Philip Durbin wrote:
Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Philip Durbin

unread,
Oct 9, 2017, 11:16:27 PM10/9/17
to dataverse...@googlegroups.com
Yeah, I'm sure it's difficult. It would be best for you to do a fresh reinstall: http://guides.dataverse.org/en/4.8/installation/installation-main.html#fresh-reinstall

Janet
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsubscribe...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/f19a576a-6d69-4fbe-90ae-3b8227583800%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Janet McDougall - Australian Data Archive

unread,
Oct 10, 2017, 6:42:54 AM10/10/17
to Dataverse Users Community
Hi Phil
A fresh install as a solution is dramatic. Is there anyone who can tell us how to reverse out of a custom metadata block?
Thanks
Janet

Pete Meyer

unread,
Oct 10, 2017, 9:01:30 AM10/10/17
to Dataverse Users Community
Hi Janet,

If you're comfortable directly editing the database, I should be able to track down the SQL commands I've used for this in the past.  Deleting custom blocks can cause undefined behavior (especially if there are any datasets using the custom blocks), and my experience doing this roughly parallels the recommendation from Phil (and others) that reinstallation is the cleanest way to go.

Best,
Pete

Janet McDougall - Australian Data Archive

unread,
Oct 10, 2017, 3:03:18 PM10/10/17
to Dataverse Users Community
Hi Pete
It would be good to hear what you've done to to manually delete the custom metadata blocks. We can edit the database directly as it's a dev server and I only used unpublished test datasets to look at the custom blocks - which I can delete.

This means we should really build any custom blocks prior to ingesting data and releasing a dataverse in production if a fresh install is really the only clean solution to back out. Metadata is core to the purpose of an archive so I imagine how it becomes problematic to undo. I prefer that we have experienced this now than in production!

Thanks for your response.
Janet

Pete Meyer

unread,
Oct 10, 2017, 4:53:55 PM10/10/17
to Dataverse Users Community
Hi Janet,


On Tuesday, October 10, 2017 at 3:03:18 PM UTC-4, Janet McDougall - Australian Data Archive wrote:
Hi Pete
It would be good to hear what you've done to to manually delete the custom metadata blocks. We can edit the database directly as it's a dev server and I only used unpublished test datasets to look at the custom blocks - which I can delete.

Attached (assuming various email filters don't decide to exclude SQL); feel free to let me know if you have any questions.  For future reference (aka - anyone seeing this in future searches): here there be dragons, don't use this in production, and make sure you've got backups.
 

This means we should really build any custom blocks prior to ingesting data and releasing a dataverse in production if a fresh install is really the only clean solution to back out.  Metadata is core to the purpose of an archive so I imagine how it becomes problematic to undo. I prefer that we have experienced this now than in production!


When developing a new metadata block, the approach I took was to use a development VM that I could wipe and re-provision as needed.  For production, you obviously wouldn't want to do that - but it seems to me that deleting parts of a metadata schema once something is in production is the kind of thing that could run into problems (aka - datasets created/published based on parts of the schema that aren't there anymore).

Best,
Pete
delete_meta.sql

Janet McDougall - Australian Data Archive

unread,
Oct 10, 2017, 5:43:45 PM10/10/17
to Dataverse Users Community
Hi Pete
Thanks and yes, def agree that deleting fields in a metadata block midway can only lead to trauma, besides being illogical. If new fields are required theN a migration to a new dataverse might be an option if existing studies need to include the new fields. Things to consider. I will also test with referencing blocks to a dv alias. Your advice and experiences appreciated.
Janet

Venkatachalam Kannadasan

unread,
May 24, 2018, 12:14:43 AM5/24/18
to Dataverse Users Community
Dear Pete

I am using your delete_meta.sql to delete the metadata created with the custom metadata block. I am facing a foreign key issue with the two tables when running the following commands, can you please help me with the sql to delete these entries?

delete from datasetfield where datasetfield.id in 
(
select datasetfield.id from datasetfield join datasetfieldtype on datasetfield.datasetfieldtype_id=datasetfieldtype.id
where datasetfieldtype.metadatablock_id=12
);

This sql command when run shows the following error

ERROR:  update or delete on table "datasetfield" violates foreign key constraint "fk_datasetfieldcompoundvalue_parentdatasetfield_id" on table "datasetfieldcompoundvalue"

DETAIL:  Key (id)=(11481780) is still referenced from table "datasetfieldcompoundvalue".


Likewise when running the command

delete from datasetfieldtype where metadatablock_id=12;

throws the following error

ERROR:  update or delete on table "datasetfieldtype" violates foreign key constraint "fk_datasetfield_datasetfieldtype_id" on table "datasetfield"

DETAIL:  Key (id)=(297) is still referenced from table "datasetfield".


And also when running the command

delete from metadatablock where metadatablock.id=12;

throws the following error

ERROR:  update or delete on table "metadatablock" violates foreign key constraint "fk_datasetfieldtype_metadatablock_id" on table "datasetfieldtype"

DETAIL:  Key (id)=(12) is still referenced from table "datasetfieldtype".


This I understand I can delete only after deleting from the other two tables. Any help would be appreciated. 

Regards
Venki

Philip Durbin

unread,
May 29, 2018, 10:10:17 AM5/29/18
to dataverse...@googlegroups.com
Venki I see you and Pete chatted in IRC so I hope you're all set: http://irclog.iq.harvard.edu/dataverse/2018-05-24

When I hack on metadata blocks I'm very aggressive about dropping my database but this is on my laptop with no real data so I can understand how this might not be the ideal workflow for you.

If you have any comments to add to the "Metadata: Dataverse project take ownership of documentation for creating metadata blocks" issue, please go ahead: https://github.com/IQSS/dataverse/issues/3168

Phil


--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/a8d3f768-2ebb-4a86-84c6-ad20cc72d869%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Philip Durbin

unread,
Oct 30, 2018, 8:14:25 AM10/30/18
to dataverse...@googlegroups.com
Hi everybody, I'm warming up this old thread because there's been some movement on documenting how custom metadata blocks.

Issue #3168 mentioned above was addressed in Dataverse 4.9.3 resulting in this "Metadata Customization" page in the Admin Guide: http://guides.dataverse.org/en/4.9.4/admin/metadatacustomization.html

It's a great starting point and I want to thank Tim DiLauro again for working on it.

I'm now working on https://github.com/IQSS/dataverse/issues/5205 which is about further improving the doc above to explain what curl commands to run, etc.

If anyone reading this would like me to document anything in particular about metadata blocks while I'm in there, please let me know either here or as a comment in the issue.

Thanks!

Phil

To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/a8d3f768-2ebb-4a86-84c6-ad20cc72d869%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--

Sherry Lake

unread,
Oct 30, 2018, 11:29:04 AM10/30/18
to Dataverse Users Community
Hi Phil,

I would love to review and contribute. I'm at a workshop currently and can look at things later this week.

--
Sherry
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

Philip Durbin

unread,
Oct 30, 2018, 2:04:45 PM10/30/18
to dataverse...@googlegroups.com
Great! Thanks! You and others are welcome to look at the proposed improvements I made to that "Metadata Customization" page: https://github.com/IQSS/dataverse/blob/567fb96e9bc0266a3bb4c6212cfc8e93b0106ce9/doc/sphinx-guides/source/admin/metadatacustomization.rst

Here's the pull request in case anyone would like to leave a review: https://github.com/IQSS/dataverse/pull/5251

Thanks,

Phil

To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/c0a7e53a-cb32-4c57-a751-2c7ff3041c01%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Philipp at UiT

unread,
Nov 1, 2018, 3:28:04 AM11/1/18
to Dataverse Users Community
Thanks, Phil. Are there any plans to make metadata blocks customizable through GUI?

Best,
Philipp
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

Philip Durbin

unread,
Nov 1, 2018, 7:05:51 AM11/1/18
to dataverse...@googlegroups.com
No, but you could create an issue about this.

To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.

To post to this group, send email to dataverse...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/5c711bbb-35df-43e0-a663-766b355003b9%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Youn Noh

unread,
Feb 24, 2025, 10:29:11 AMFeb 24
to Dataverse Users Community
Is there any reason not to allow ISO format dates with millisecond precision as allowed by Solr [1]? What would happen if one attempted to add or update metadata using the Native API and a date field contained an ISO format date? For now, if we are using the Native API, should we truncate to YYYY-MM-DD? Thanks.

Philip Durbin

unread,
Feb 24, 2025, 10:32:28 AMFeb 24
to dataverse...@googlegroups.com
You can try it and see, but I'm pretty sure the validation will complain if it's not in YYYY-MM-DD format.

You are very welcome to open an issue about more granular dates and times.

Solr is an implementation detail we probably don't want to tightly couple Dataverse to but it's a good reference for the sort of timestamps other systems accept.

I hope this helps!

Phil
To view this discussion visit https://groups.google.com/d/msgid/dataverse-community/0f95d959-9706-4295-af94-d53a726701b2n%40googlegroups.com.

Clemens, Vera

unread,
Feb 24, 2025, 10:51:27 AMFeb 24
to dataverse...@googlegroups.com
Hi,

chiming in because I've previously had a look at the date validator code and found that it allows some formats which are not currently documented:

* YYYY followed by AD or BC
* a "Bracket format" (not familiar with this)
* datetime formats (yyyy-MM-dd'T'HH:mm:ss, yyyy-MM-dd'T'HH:mm:ss.SSS and yyyy-MM-dd HH:mm:ss).


So, I think adding or updating metadata with any of these formats should succeed.

Note that one comment in the code does state that this is a "temporary fix", so I guess this might be subject to change :)

Best regards,

Vera

Youn Noh

unread,
Feb 24, 2025, 11:04:31 AMFeb 24
to Dataverse Users Community
Thanks, Phil and Vera. Alternatively, if a value in a date field has a lower precision than assigned in defining the field, will it cause errors (e.g., YYYY value in a field defined as YYYY-MM-DD)? We could try different options, but if anyone has prior experience, that would be helpful. Thanks.
Reply all
Reply to author
Forward
0 new messages