[Dataset quality] Show datasets namespaces

[yngrdyn]

:notebook: Summary

Allow users to be able to inspect namespaces within datasets. For this convert datasets with different namespaces in expandable rows.

• :art: Figma: https://www.figma.com/file/aaJNMsOklYhjwtu5TP7UDK/%5BLogs%2B%5D-Organize-and-Admin?node-id=2106%3A381912&mode=dev

:heavy_check_mark: Acceptance criteria

• Datasets with namespaces are expandable rows, these rows don't have actions associated, and each child of the row has the following actions
  • Shows Size, Malformed docs, Last activity.
  • Actions are targeting directly the namespace in the row, for example navigating to Log explorer redirects the user to Log explorer with the namespace selected.
• Dataset with namespaces are showing a summation on all the stats for their datasets in the respective columns.
• Datasets without namespaces are not expandable.

:bulb: Implementation hints

• For the redirection to Log explorer we have different approaches, for example we could use KQL, we could use Unified Search pills or ultimately we could use Namespace selector. For the last option we should be sure namespace is always a filter defined for the integration so we could add this property to the locators of the log explorer, this approach would imply more work than the other two approaches.

[elasticmachine]

Pinging @elastic/obs-ux-logs-team (Team:obs-ux-logs)

[mohamedhamed-ahmed]

Are we going to have the Actions enabled on dataset level also? or only on namespace level?

[gbamparop]

@isaclfreire are there any unknowns or blockers from the UX side for this work?

[isaclfreire]

Not really a blocker but this could be an opportunity to review this list design and explore if it makes sense to align it with the designs made for the Datasource selector that has a very similar need (formatting the list by Integration + inner datasets, whereas here would be by Integration + inner namespaces). This is just food for thought, not a dependency.