Assess the correct use of ConcreteReadAttributePath in the read call-path for the IM

[mrjerryjohns]

This PR introduces a ConcreteReadAttributePath that contains a list index. However, it adds the use of that fairly late in the call chain for handling a read (inside the ReadSingleClusterData function). We should assess where the right place for this to be is.

[mrjerryjohns]

Assigning this to you @erjiaqing since you're gonna be in this space a fair bit.

[bzbarsky-apple]

Fwiw, I think ReadSingleClusterData should definitely take a ConcreteReadAttributePath.

[bzbarsky-apple]

#12067 fixes the obvious thing. Need to figure out how to share the path-expanding iterator between read and write paths before we can push things further.

[holbrookt]

@bzbarsky-apple @mrjerryjohns is this needed for 1.0? If not, please remove the label. Thanks!

[bzbarsky-apple]

I don't think the remaining work here is a 1.0 blocker, personally.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.

[mrjerryjohns]

Already done.