When rendering Markdown preview, Unicode newline characters U+2028 LINE SEPARATOR and U+2029 PARAGRAPH SEPARATOR are replaced with an empty string at the source preprocessing stage.

This behavior was introduced in #63936 to fix #63749, but it breaks the document in multiple ways:

• double-click word selection does not work as expected,
• screen readers may announce the text incorrectly,
• line wrapping and hyphenation does not work correctly,
• markdown-it plugins that rely on \b or \s may work incorrectly,
• even linkify fails:

  https://foo.bar/baz<LS>Lorem ipsum
  (Expected link: https://foo.bar/baz)
  (Received link: https://foo.bar/bazLorem)
  

• and so on.

Since then it was fixed at the markdown-it side, and #63936 is no longer needed:

• markdown-it/markdown-it@d9cb3cc
• markdown-it/markdown-it@faecae0
• https://github.com/markdown-it/markdown-it/blob/9.1.0/lib/rules_core/normalize.js

It would be better to revert this change, and keep the Markdown source as is before feeding it to the parser.

Steps to reproduce:

1. Create a Markdown document like this:

   https://foo.bar/baz
Lorem ipsum. Lorem
ipsum.
   https://foo.bar/baz
Lorem ipsum. Lorem
ipsum.

2. Open preview:
   • Expected result: LS and PS are displayed as is, links end at Unicode newlines.
   • Actual result: no Unicode newline characters displayed, links also consume the immediately following word.

Does this issue occur when all extensions are disabled?: Yes

• VS Code Version:

  Version: 1.73.0 (system setup)
  Commit: 8fa188b2b301d36553cbc9ce1b0a146ccb93351f
  Date: 2022-11-01T15:34:06.111Z
  Electron: 19.0.17
  Chromium: 102.0.5005.167
  Node.js: 16.14.2
  V8: 10.2.154.15-electron.0
  OS: Windows_NT x64 10.0.22000
  Sandboxed: No
  

• OS Version: Windows_NT x64 10.0.22000