Update pane index when moving panes in same tab

[scauligi]

Fixes #6028.
Not sure if this is the best solution, but it works.

[wez]

Thanks for this!

I have a suggestion that I think makes the flow clearer overall, and which will be more robust against other changes that might mutate the tab contents in the future

[scauligi]

I think this change is in line with what you're suggesting? There's a minor complication in that the pre-removal pane_index is needed to calculate split_size, but the post-removal pane_index is what we need for the final tab.split_and_insert call.

[wez]

Yeah, this is what I was thinking. The complication you mentioned just sounds like how this works now, or did you mean that phrasing the code this way introduces another issue?

[scauligi]

Whoops! Lost track of this and forgot to reply.
The "complication" was in reference to the fact that the new pane index is needed for the tab.split_and_insert call, so recalculating it inside the nested scope wouldn't work since we need the new index in the outer scope (which my latest change should be taking care of).