Revert "Delay list -> dict conversation to end (#192)"

[blink1073]

This reverts commit 7b9c024.

Temporary fix for jupyterlab/jupyterlab#11743

[codecov-commenter]

Codecov Report

❗ No coverage uploaded for pull request base (main@cce8b4e). Click here to learn what that means.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main     #234   +/-   ##
=======================================
  Coverage        ?   76.08%           
=======================================
  Files           ?       29           
  Lines           ?     2204           
  Branches        ?        0           
=======================================
  Hits            ?     1677           
  Misses          ?      527           
  Partials        ?        0           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update cce8b4e...3bcbb93. Read the comment docs.

[fcollonval]

I think we can reapply #192 now that jupyterlab/jupyterlab#11744 is merged?
cc @jtpio

[jtpio]

It's not clear whether #192 should be considered as a bug fix or not. If so and assuming get_page_config is not really part of the public API then we could reapply it as a bug fix, now that jupyterlab/jupyterlab#11744 handles both cases.

[fcollonval]

Ok so it would be good to wait for a feedback of @vidartf before taking a decision.

I added a note to bring that up at the next meeting.

[vidartf]

Sorry, I was out for the holidays. Did that change break anything more than the test? get_page_config only gives the config from disk, and not the "extra" settings that the handler adds, so it has a somewhat limited scope already. The issue is ofc that it modifies the type of some returned values. Since the dict -> list conversion is mainly intended to be for the benefit of the frontend, I would argue that it makes sense to only apply this transform right before it sends the data.

#192 can be considered more of an enhancement, but it can be argued as a bug fix in that the values in page_config_data are not applied as you would expect.

[jtpio]

Did that change break anything more than the test?

Looks like it affected the disabling and enabling of JupyterLab extensions, based on this comment: jupyterlab/jupyterlab#11744 (comment)

#192 can be considered more of an enhancement, but it can be argued as a bug fix in that the values in page_config_data are not applied as you would expect.

So we could maybe reapply #192? Especially now that jupyterlab/jupyterlab#11744 should handle both cases?

[vidartf]

So we could maybe reapply #192?

Probably, but it might not make sense as a patch/minor release if it breaks older versions of lab.