fix(NcRichText) - don't strip <tag>-like content when parsing markdown

[Antreesy]

☑️ Resolves

• Replace legacy package https://github.com/remarkjs/remark-external-links/releases/tag/9.0.1
• Fix NcAvatar render for shown initials, when use different variant of white-space rule (common for rich components)
• Ease HTML tree for markdown parsed content:
  • before each child node (also text node) was wrapped with span before inserted into parent node
  • now they are inserted directly
  • Revert some previous changes to restore behavior back to v7.12.2 and older (forced new lines):
    • fix(NcRichText) - keep newlines in rendered Markdown  #4461
    • fix(NcRichText) - don't populate a first newline to properly parse blockqoute tags #4470
  • Apply some base styling to markdown rendered content:
    • margin-bottom for separate text blocks
    • bold headings
    • aligned lists
    • stylised blockquotes
• Escape/unescape special symbols < and > before parsing:
  • Leading < could be treated like a XML tag start and be stripped with inner content (because of possible XSS attack)
  • Leading > should be unescaped to correctly parse blockquotes
• Fix NcRichContenteditable pasting HTML alike content strips off text #3908
• Fix Markdown rendering "eats" html tags nextcloud/spreed#10412

🖼️ Screenshots

🏚️ Before	🏡 After
commits 71c2cbd, 836b7c8	No visual changes
commit 2e457f3	NcAvatar fix
commit 62fa807	ease HTML fix
commit 22d0c6f	Revert old commits, add base styling
commit 38606f2	Escape XML when parsing markdown

🚧 Tasks

• Code review
• Visual check
• Security check

🏁 Checklist

• ⛑️ Tests are included or are not applicable
• 📘 Component documentation has been extended, updated or is not applicable

[DorraJaouad]

Tested and it works.
It solves the issue of skipping HTML-like content though there might be a better solution as follow-up .

[Antreesy]

Rebased onto master, split commits with documentation to different PR #4507

[ShGKme]

Works weird with multiline new lines in quotes

Text	Render

[Antreesy]

Rebased onto master with some changes

This returns a behavior, then Markdown parser eats newlines (Enter, Shift+Enter), but it seems to be a default behavior for parsers (see how Github also eats several newlines - edit this comment)