Function ToText

internal/htmlutil/htmlutil.go:12–25

ToText converts HTML content to plain text, preserving basic structure.

(s string)

// ToText converts HTML content to plain text, preserving basic structure.
func ToText(s string) string {
	doc, err := html.Parse(strings.NewReader(s))
	if err != nil {
		return s
	}
	var b strings.Builder
	walkNode(&b, doc)
	// Collapse runs of 3+ newlines into 2
	result := b.String()
	for strings.Contains(result, "\n\n\n") {
		result = strings.ReplaceAll(result, "\n\n\n", "\n\n")
	}
	return strings.TrimSpace(result)
}

// ExtractImageURLs finds image URLs from <img src> tags and
// <figure data-trix-attachment> elements.