refactor(rich-markdown-editor): hoist paste-strip regex, cross-ref block-mover

/simplify pass follow-ups: hoist the <style>/<script> strip regex to module scope (matches
the file's other module-level pattern consts, avoids re-alloc per paste), and note the
Mod-Shift-Arrow block-reorder chords in the keymap doc so arrow handling is discoverable
across the two files. No behavior change.

Author: waleedlatif1

apps/sim/app/workspace/[workspaceId]/files/components/file-viewer/rich-markdown-editor/keymap.ts

@@ -128,7 +128,8 @@ function selectAdjacentSelectedLeaf(editor: Editor, direction: 'up' | 'down'): b
  *   same scoped behavior as a code editor.
  * - **ArrowUp/ArrowDown** select an adjacent divider or image, whether arrowing off a textblock edge
  *   ({@link selectAdjacentLeaf}) or stepping from one already-selected leaf to the next
- *   ({@link selectAdjacentSelectedLeaf}).
+ *   ({@link selectAdjacentSelectedLeaf}). (The `Mod-Shift-Arrow` block-reorder chords live separately
+ *   in `./block-mover.ts`.)
  *
  * Plus a plugin that (a) highlights dividers/images falling inside a range selection (e.g. select-all),
  * which the browser's native text highlight skips because leaves carry no text, and (b) flags the

apps/sim/app/workspace/[workspaceId]/files/components/file-viewer/rich-markdown-editor/markdown-paste.ts

@@ -62,14 +62,17 @@ function parseVscodeLanguage(data: string | undefined): string {
   }
 }
 
+/** `<style>`/`<script>` elements (with their content), matched as a pair via the tag backreference. */
+const NON_CONTENT_HTML = /<(style|script)\b[\s\S]*?<\/\1>/gi
+
 /**
  * Strips `<style>`/`<script>` elements from pasted HTML. Google Sheets and Word prepend a `<style>`
  * block of CSS (and Sheets a `<google-sheets-html-origin>` wrapper); ProseMirror's DOM parser has no
  * rule for `<style>`, so it would walk the element's CSS text into the document as literal paragraphs.
  * Removing these before parsing keeps the pasted content clean (PM already discards unknown wrappers).
  */
 function stripNonContentHtml(html: string): string {
-  return html.replace(/<(style|script)\b[\s\S]*?<\/\1>/gi, '')
+  return html.replace(NON_CONTENT_HTML, '')
 }
 
 /**