Previously, the Editor component used character-level (grapheme-level)
wrapping, which broke words mid-character at line boundaries. This
created an ugly visual experience when typing or pasting long text.
Now the Editor uses word-aware wrapping:
- Wraps at word boundaries when possible
- Falls back to character-level wrapping for tokens wider than the
available width (e.g., long URLs)
- Strips leading whitespace at line starts
- Preserves multiple spaces within lines
Added wordWrapLine() helper function that tokenizes text into words
and whitespace runs, then builds chunks that fit within the specified
width while respecting word boundaries.
Also updated buildVisualLineMap() to use the same word wrapping logic
for consistent cursor navigation.
Added tests for:
- Word boundary wrapping
- Leading whitespace stripping
- Long token (URL) character-level fallback
- Multiple space preservation
- Edge cases (empty string, exact fit)
- Export wrapTextWithAnsi from tui package
- Add 'Handling Input' section to README with key detection example
- Document wrapTextWithAnsi in utilities section
