feat: synthetic data generation cascade down topic tree

[leonardmq]

What does this PR do?

• Add Add to all buttons on the Synthetic data page to let the user generate samples for every subtopic leaf descendant of a given topic node; I called that "Cascade mode"
• Add optional parallelization of synthetic data generation when in Cascade mode

The parallelization is inspired by what was done for the Save all action; the error handling is similar. Tried to keep the workers' errors mapped to their topic so that if we want, we can later add something to Try again and retry only the failed topics.

Related Issues

N/A

Contributor License Agreement

I, @leonardmq, confirm that I have read and agree to the Contributors License Agreement.

Checklists

• Tests have been run locally and passed
• New tests have been added to any work in /lib

[leonardmq]

Was there a plan to skip the duplicates?

[leonardmq]

Wondering if there’s a more idiomatic way in Svelte to control modal visibility (when the modal decides to close itself) than passing in a callback from the parent component like this

[leonardmq]

The Add data to all button is added on the root of the tree as well as on each topic (that has 1 or more child topics):

The Add data to all button is hidden on topics that have no children:

The Add data to all modal is the same as Add data with the text slightly altered to align with what it does, and adds a parallelization field:

If the generation is successful, it will close the modal.

If the generation runs into errors, the errors from each worker are collected and displayed at the end (specific error message a little clunky - might want to rephrase):

In the case of a non-cascade (i.e. single topic) generation failure, the error is shown directly like this: