test: Real browser tests

[reidbarber]

Adds setup for using Vitest Browser Mode to run certain tests in real browsers (Chromium, Firefox, and WebKit).

Tests with the *.browser.test.{ts,tsx} pattern get run when running yarn test:browser.

Notes:

• This currently doesn't run in CI, but we can enable that later if we want. These take a while to run, so we probably want to limit them to testing things that are difficult to test in our existing tests without mocking browser features that aren't available in jsdom. We can still rely on Chromatic for basic browser rendering.
• This PR adds a single DropZone.browser.test.tsx to get us started. Unfortunately, I noticed our drag and drop events don't work with userEvent.dragAndDrop in Chromium because of how Playwright interfaces with Chromium, so I made a utility that falls back to manually firing the DragEvents in the expected order. Hopefully we don't have too many things like this to custom handle.

✅ Pull Request Checklist:

• Included link to corresponding React Spectrum GitHub Issue.
• Added/updated unit tests and storybook for this change (for new code or code which already has tests).
• Filled out test instructions.
• Updated documentation (if it already exists for this component).
• Looked at the Accessibility Practices for this feature - Aria Practices

📝 Test Instructions:

Run yarn test:browser or yarn test:browser run to run without watch mode.

🧢 Your Project:

[rspbot]

Build successful! 🎉

• View the storybook
• View the S2 storybook
• View the S2 docs
• View the React Aria docs

[rspbot]

Build successful! 🎉

• View the storybook
• View the S2 storybook
• View the S2 docs
• View the React Aria docs

[LFDanLu]

Happy to merge it in as is, we can discuss how specifics like run cadence, CI, and what tests to use this for after playing around with more test scenarios. Also curious if this easily extends to the more classic kind of browser tests where we send it to a page/build instead of calling render on our own written test case

[LFDanLu]

looks like we need to run yarn playwright install to download chromium and other browsers before this works. I assume we'd want to be as up to date as possible (and thus download fresh on every run?) but track what versions failed when that happens for browser specific bug logging. Alternatively we could store the browser builds somewhere I'd imagine and update only when desired

[LFDanLu]

a bit unfortunate that we still have to mock/dispatch events here

[LFDanLu]

I noticed that the tests were being run at a mobile device screen resolution, would be nice to have it run in desktop widths as well automatically. Also I assume there is the ability to run headless as well?