Repository navigation
Test and Validate DCAT-US3 dataset harvest #6177
Description
In order to validate our status on DCAT-US3 harvest support, we want to test our harvester in either stage or prod and validate end to end working example that is searchable and UI supported.
This involves having a valid DCAT-US3 catalog (for testing purposes) on the web, and validating the successful harvest source creation, harvest job, and appearance of the dataset(s) from the example on catalog-stage.data.gov.
Any issues/problems should be documented and addressed as follow up stories.
Metadata
Metadata
Assignees
Type
Projects
- Status
Activity
SueValente commentedon Jul 29, 2026
@jbrown-xentity can you add what it would look like to satisfy this issue?
jbrown-xentity commentedon Aug 3, 2026
Updated the description with more clarity.
Looks like there are a few minor UI tweaks/bugs due to DCAT-US3 changes, but everything is working end to end for DCAT-US3 specifically for the Dataset class.
I created a harvest source here. This resulted in 2 datasets, Employment Statistics and Annual Budget Data. This can be compared against a normal GSA record. I think the panels on the right being blank is fine if they don't have information, such as Access & Use or Dataset Information. Because the source datasets didn't have access use or rights, or modified field, this information is blank. There are also no distributions, so that part is correct and is excluded already.
One final tweak I would like to make is the "Find Related Datasets" section. Since keywords was required in 1.1 this area was always populated, either with keywords or with keywords and link to
isPartOfrelated dataset. With 3.0 it's not required, and I would like that to be excluded if nothing is added there.This would involve simply combining the 2 checks here and here, and if neither are positive then excluding the whole section here. Going to spin up a quick ticket to that effect, as it should involve a quick test as well.
As that is a minor cosmetic annoyance, we can consider this ticket complete. We have proved that the DCAT-US3 harvest is working as expected today to harvest datasets all the way into catalog. The epic still needs to capture some new classes, but we are on our way.