fix: propagate hidden flag to computeAccessibleName

[foxaltus]

What: this PR fixes #820 (comment)

Why: *ByRole APIs are slow, because the DOM environment relies on getComputedStyles to compute elements visibility. This can be mitigated by using the { hidden: true } option. However, when using *ByRole APIs along with the name option, the performance toll is still being paid because computeAccessibleName is called internally, which does not currently ignore hidden elements.

How: this PR propagates the hidden option to computeAccessibleName, which avoids paying the performance toll.

Checklist:

• Documentation added to the
  docs site
• Tests
• TypeScript definitions updated
• Ready to be merged
  

[codesandbox-ci]

This pull request is automatically built and testable in CodeSandbox.

To see build info of the built libraries, click here or the icon next to each commit SHA.

Latest deployment of this branch, based on commit 9bc828e:

Sandbox	Source
react-testing-library-examples	Configuration

[foxaltus]

Hmm I realize this is not enough...

The changes allow to bypass the expensive isHidden check at https://github.com/eps1lon/dom-accessibility-api/blob/main/sources/accessible-name-and-description.ts#L601, but not the ones at https://github.com/eps1lon/dom-accessibility-api/blob/main/sources/accessible-name-and-description.ts#L388. So the toll is still being paid :(

See: most time is being spent in computeMiscTextAlternative:

(FYI this test looks by table rows by name in a 25 rows x 10 cols table)

[github-actions]

Uh oh! @foxaltus, the image you shared is missing helpful alt text. Check #1371 (comment).

Alt text is an invisible description that helps screen readers describe images to blind or low-vision users. If you are using markdown to display images, add your alt text inside the brackets of the markdown image.

Learn more about alt text at Basic writing and formatting syntax: images on GitHub Docs.

[foxaltus]

At this point I think that maybe testing-library should offer an option to fake getComputedStyle (similar to the computedStyleSupportsPseudoElements option). In some of my tests (dealing with tables) this improves the performance by a factor of 10.

If this sounds too much of a bazooka solution, maybe we should instead consider adding an option at *ByRole level, such as namingStrategy?: 'fast' | 'accurate':

• fast would send a mock getComputedStyle implementation to computeAccessibleName.
• accurate (default) would use the current implementation (+ I think it could honour the hidden option, as already done in this PR)

Thoughts?