E2E Utils: Wait for canvas instead of window.wp

[WunderBart]

What?

In testing against a slower env (wp.com in my case) it turned out that the window.wp.* is ready before the canvas is, so calling utils like insertBlock immediately after creating a post has no effect. Waiting for the canvas instead gives stable results and makes the affected utils work as expected.

Why?

For stability.

Testing Instructions

Nothing particular. All the tests should pass.

[github-actions]

Warning: Type of PR label mismatch

To merge this PR, it requires exactly 1 label indicating the type of PR. Other labels are optional and not being checked here.

• Type-related labels to choose from: [Type] Automated Testing, [Type] Breaking Change, [Type] Bug, [Type] Build Tooling, [Type] Code Quality, [Type] Copy, [Type] Developer Documentation, [Type] Enhancement, [Type] Experimental, [Type] Feature, [Type] New API, [Type] Task, [Type] Technical Prototype, [Type] Performance, [Type] Project Management, [Type] Regression, [Type] Security, [Type] WP Core Ticket, Backport from WordPress Core.
• Labels found: [Package] E2E Test Utils.

Read more about Type labels in Gutenberg. Don't worry if you don't have the required permissions to add labels; the PR reviewer should be able to help with the task.

[WunderBart]

This one's breaking because the editor.canvas locator does not automatically switch when tests switch to legacy canvas. I know this was discussed before, but maybe some new ideas have come up since? @kevin940726 @Mamaduka

I've refreshed my research, and it still doesn't seem to be supported natively by Playwright, as the locator.or cannot be mixed with a frame locator, e.g.

canvas() {
  return this.page.locator('.canvas').or(
    this.page.frameLocator('canvas-iframe').locator('.canvas')
  );
}

The only way that would work would be making the canvas getter async, which would break almost all the tests as it would need to be awaited. 😅