Dataframe view: add an Index column for instances and support un/collapsing

[abey79]

• Un/collapsing by clicking on the +
• Collapsing by clicking on the vertical line

[abey79]

@gavrelina I like the look of the [+] / [–] thing, but shouldn't we be using the same triangle as everywhere else?

[gavrelina]

@gavrelina I like the look of the [+] / [–] thing, but shouldn't we be using the same triangle as everywhere else?

You're right.
Using ▶ and ▼ would make it more consistent.

Initially [+] [-] was there to prevent repeating similar icons for 'sorting' and possible similarities with the 'static/temporal' components. But neither of those is used now (!! Except the sorting icon for Chunk store browser, which we shall also change to ), hence it's not justified.

Screenshot with ▶ and ▼ for visibility:

[gavrelina]

New proposal:

Components are grouped by entities (before we can group them by archetypes through tagged components)

• Each of the instance could be "uncollapsed" until the last index of its entity.
• One can uncollapse by pressing an icon, which appears only on hover
• Simultaneously the other instances are uncollapsed as well until the index of the initiated item. I.e. compA[0-5], compB[0-15], if I uncollapse via compA I will see CompA[0,1,2,3,4,5] and for CompB[0,1,2,3,4,5]+10 more.
• If there are more instances for certain component we display "N more" with ability to uncollapse further.
• When uncollapsed — each of the rows serves as a way to collapse everything back to a "summary row".
• When uncollapsed : we still see a summary row with the total number of instances for each component.
• If one scrolls down to uncollapsed instance batch that is longer then the rest -> we explore the sticky behavior for the latest known values for other components.
• have a separate column with (shared) indices => to help users see which [N] is a certain value they are looking at instead of calculating (is it [61?] or [62?]).

Interactions:
https://github.com/user-attachments/assets/2c6ef858-9aff-4181-ad96-d7fe15504adc

[gavrelina]

@abey79
one more improvement thingy!
to add "zebra" lines for the ones that are expanded (to make it easier to scan through them)? I believe @emilk mentioned it was possible!

[abey79]

@gavrelina I'll do that in a follow-up PR, should be too back. Tracking this here:

• Have alternating background stripes also in expanded rows in the dataframe view #7412