Rich and localized text API

[bartfeenstra]

• document-based (this would actually be a separate follow-up feature)
  • titles
  • dates
  • authors
• per-text content type, e.g. any piece of natural language has a content type and a locale
• allow plain text extraction (for search index)
• Implement this for Note and Described
• Introduce a new RichText interface to capture this
• Update existing methods to return not just str but str | RichTextStr where RichTextStr(RichText, str)
• Can this leverage renders somehow? Maybe those should provide extensions and content types they can render. Or better, just content types and keep a global content type map. Does a library for that already exist?
• add Renderer.render()? Separate this out into a dedicated PR.
• we'd need an HTML 'filter' Renderer to allow specific HTML only. This filter MUST NOT be used with other renderer such as Jinja2!
• allow Gramps notes to get an attribute specifying their content type? If given, do not parse markup XML. If not given, parse markup XML to HTML and set the note as text/html

An example of how Gramps exports notes:

  <notes>
    <note handle="_fb02f4ee0eb2e4ac087c5d16b99" change="1724938235" id="N0000" type="Person Note">
      <text>This is a note WITH BOLD TEXT and a link?</text>
      <style name="bold">
        <range start="15" end="29"/>
      </style>
      <style name="link" value="gramps://Person/handle/f99770ed37f7eb61d0b2c92f0a1">
        <range start="34" end="40"/>
      </style>
    </note>
  </notes>

This can be used to represent pages with managed content, but also notes attached to entities