Consider LTR embedding marks for display in the a11y view.

[samreid]

From phetsims/joist#994, one idea was to Consider LTR embedding marks for display in the a11y view. This may become more important as we start to add RTL language translations in the a11y key.

[jbphet]

There are a couple of things that I'd like to note for this issue:

• I recently did a search on whether directionality characters could impact what screen readers say, and the response from Google Gemini was this:

  Yes, Unicode control characters that manage text directionality can affect screen readers, and how they do so varies by device and software. In most cases, these characters are not announced by the screen reader but influence how it reads the text, which is typically beneficial for users who require bidirectional content.
  So we should be careful that whatever we do for the a11y view doesn't negatively impact screen readers for these languages. I'll be the first to admit that this may be a hard thing to test, but we should be aware of it.

• The title of this issue refers to "LTR embedding marks", but we should probably be considering all directional control characters (also called "bidirectional" or "bidi" control characters), including isolates and direction control characters.
• There has been some work done lately on improving our handling of RTL languages, specifically for mixed-direction strings. That work is currently on hold, but the learnings there should inform what is done here. The current "master issue" for this is Improve our handling of technical right-to-left strings in sim translations rosetta#475.
• As part of the recent work on better handling of RTL strings, a screen was added to Chains with a bunch of RTL patterns that we are using for testing. Maybe Chains could also serve as a place to test out RTL behavior for a11y strings.