Please provide a warning for each inheritance case, and see my comment for studyParClean, otherwise I'm fine.

OK. I fixed the studyParClean. In a bit different way - I keep studyPar as it was without changing it. And I only clean the fields from the result. This is even more clean solution.

I made a few more edits addressing a few more comments. Can I resolve the rest?

Regarding the warnings - we have to keep it relatively simple. And we can't report on general basis. As we want to keep things modular and this report will change based on what subjects are included and can therefore differ on a re-run with a more complete dataset.

So I propose to report something like this:
Importing subject XX, session YY, run ZZ.
Note: fields AA, BB, CC were updated as multiple studyPars were provided in studyPar.json

Importing subject XX, session YY, run ZZ.
Exception subject 11|12, session 2|3, run1: Updated fields AA, BB, CC
Exception subject 12, session 3, run1: Updated fields AA, BB, CC, DD

Any preference?

HENK: When it reads the non-first study par, it warns "Note that for subjects XX sessions YY we're using parameter XX instead of YY"

And check the main issue comment.

I have copied the same answer to PR as well:

• We have multiple studyParameters, not sure where "context" comes from?
  Good catch. That's because we called it ImportContexts previously. I have fixed this everywhere. 501c564

• We create a function that parses studyPar.json. This provides a heads-up to the user for each aliasHierarchy, e.g.,:
  "Note that for cases that match subject regular expression ˆsub-\d{3} session regular expression ... run regular expression, we will change the data parameters XX YY and ZZ.""

This doesn't make sense IMHO. Because this only states the basic functionality of the multi-parameter studyPar :) It doesn't make to report general warning, it makes sense to only report specific warnings to cases that are actually happening.
So I would go with the warning you previously agreed upon: "When it reads the non-first study par, it warns "Note that for subjects XX sessions YY we're using parameter XX instead of YY"

An example of a current warning is:
Evaluating multi-parameter studyPar for subject/visit/session: Sub1/1/2
For RegExp subject/visit/session //2, overwriting fields: LabelingDuration, PostLabelingDelay, TimeEncodedMatrixSize, ASLContext,

• Later, we will expand this with provenance: Import provenance #1190
  OK. Will check later

I agree in general with #1193, but note that we will anyway revamp the legacy structure of parameters that dataPar uses - so things might still change a bit. Anyway - I agree that we have to clean that. These things happen step-by-step - don't criticize when things are a mess, because they were a super-mess before :D

No I wouldn't provide this information on a subject-by-subject basis, because that will be lost in the long log-list. Instead, at the beginning of the whole conversion: you give a single big warning:

Evaluating multi-parameter studyPar for all subjects/visits/sessions:
For subjects 001, 002, 003, visits 002, session 001, we overwrite the following fields: LabelingDuration, PostLabelingDelay, TimeEncodedMatrixSize, ASLContext:

• LabelingDuration 1.8 becomes 2.0
• PLD 2.0 becomes 1.5
  ..
  But I'm fine if you move this to the future provenance issue, and close this conversation.

JAN: But this is verbatim what is written in the studyPar.json...
HENK: Exactly, the warning is to help the user before most of the (potentially parallel) import starts. The studyPar.json exception thingy is not the easiest thing in the world, users like overview; so you give them a simple overview. Indeed nearly what is in the studyPar.json. But then again, it is a warning, just a kind reminder. Of course, if you're smart, you don't need this warning.

Main reason is that your warnings are lost in the gigantic list of import text. If you concentrate all studyPar.json-related warnings as I suggest, this perhaps helps more.

We can also have both.

JAN: OK. We keep the current small warning per subject. And we will try to make an overview reminder for all in a new issue.