mynode/packages/app-desktop/gui/NoteEditor/utils/cleanMarkdownNbsp.ts

/**
 * Removes standalone ` ` and related non-breaking whitespace entity lines
 * (commonly produced when converting empty rich text paragraphs `<p>&nbsp;</p>` to Markdown)
 * while safely preserving:
 * 1. Fenced code blocks (``` or ~~~) containing `&nbsp;`
 * 2. Inline `&nbsp;` within sentences or tables (e.g. `word1&nbsp;word2` or `| &nbsp; |`)
 * 3. Inline code spans (e.g. `` `&nbsp;` ``)
 *
 * @param text The input markdown text.
 * @returns Cleaned markdown text without raw standalone `&nbsp;` lines.
 */
export function cleanMarkdownNbsp(text: string): string {
	if (!text || typeof text !== 'string') return text;
	if (!text.includes('&nbsp;') && !text.includes('&#160;') && !text.includes('&#xA0;')) return text;

	const lines = text.split(/\r?\n/);
	const eol = text.includes('\r\n') ? '\r\n' : '\n';
	const result: string[] = [];
	let inCodeBlock = false;

	// Matches lines consisting exclusively of one or more non-breaking space entities
	// (&nbsp;, &#160;, &#xA0;) and whitespace / unicode non-breaking spaces (\u00A0)
	const standaloneNbspLineRegex = /^[ \t\u00A0]*(?:(?:&nbsp;|&#160;|&#xA0;)[ \t\u00A0]*)+$/;

	for (const line of lines) {
		const trimmed = line.trim();
		if (trimmed.startsWith('```') || trimmed.startsWith('~~~')) {
			inCodeBlock = !inCodeBlock;
			result.push(line);
			continue;
		}

		if (inCodeBlock) {
			result.push(line);
			continue;
		}

		// Check if the line consists entirely of standalone non-breaking space entities
		if (standaloneNbspLineRegex.test(line)) {
			// If previous line was already empty, skip pushing another empty line
			// to avoid ballooning vertical whitespace
			if (result.length > 0 && result[result.length - 1] === '') {
				continue;
			}
			result.push('');
			continue;
		}

		result.push(line);
	}

	return result.join(eol).replace(/(\r?\n){3,}/g, '$1$1');
}

export default cleanMarkdownNbsp;