[Table] Migrate TableBody to emotion

[natac13]

• I have followed (at least) the PR section of the contributing guide.

#24405

[mui-pr-bot]

@material-ui/core: parsed: +0.08% , gzip: +0.06%

Details of bundle changes

Generated by 🚫 dangerJS against 480c2f8

[mnajdova]

For fixing the error - https://app.circleci.com/pipelines/github/mui-org/material-ui/37282/workflows/f4054c2a-44af-4550-8ada-1e1c75514d81/jobs/220749

I believe we need to introduce a new option to the describeConformanceV5, so that we can wrap the element in <table /> and the structure would be correct. We can use something similar to the mount option, or introduce a callback of how to find the element, for example node.firstChild.firstChild and invoke the describeConformanceV5 with describeConformance(renderInTable(<TableBody />), ....).

[natac13]

Alright I have taken a shot at getting conformanceV5 working with a wrapper function.
I do not think the way I have done this is the best and I am very open to suggestions.
I am still failing the theme variant props test.

[mnajdova]

Alright I have taken a shot at getting conformanceV5 working with a wrapper function.
I do not think the way I have done this is the best and I am very open to suggestions.
I am still failing the theme variant props test.

Did you try to introduce some render option, similar to how the mount is defined? I think that would require less changes in the tests themselves. If the option is not provided, we can default to the render we are using at this moment. cc @oliviertassinari for thoughts on this

[natac13]

Oh I see what you mean.
Provide a render to wrap in table but use the normal render when not.
I'll make the changes

[mnajdova]

Oh I see what you mean.
Provide a render to wrap in table but use the normal render when not.
I'll make the changes

Exactlly, just watch out that the return result from render is a bit different, I suppose you would need to alter the container from the result object.

[natac13]

Ok let me know what you think of 370b4c5

Maybe once we get this figured out I should wait till it gets merged to next, then rebase next for the TableCell PR?

[mnajdova]

Ok let me know what you think of 370b4c5

Maybe once we get this figured out I should wait till it gets merged to next, then rebase next for the TableCell PR?

Exactly what I had in mind, I would maybe just suggest renaming the param to render, but let's wait on @oliviertassinari or @eps1lon feedback before doing any other changes. Also would propose moving these changes in a separate PR

[natac13]

Also would propose moving these changes in a separate PR

I was just wondering the same thing as I re-read my comment

Ill wait to hear the instructions.

Ill Start on a none table component until we get this merged into next

[oliviertassinari]

Agree with Marija, render sounds better and more consistent with what Sebastian did with mount.

Suggested change

	wrapRender: (node) => {
	render: (node) => {

[natac13]

Am I still ok to use wrapRender in the conformance tests as the name would conflict with the normal render?

[eps1lon]

Marking this for future follow-up: check-out something like render(node, { container: document.createElement('tabl') }) (need to check for connectedness and cleanup)

[natac13]

I suck at rebasing... Ill open a new PR