slow_wild_backend/internal/utils/html.go

package utils

import (
	"regexp"
	"strings"
)

// StripHTML 去除HTML标签，保留纯文本
func StripHTML(html string) string {
	// 移除HTML标签
	re := regexp.MustCompile(`<[^>]*>`)
	text := re.ReplaceAllString(html, "")

	// 替换多个空格为一个空格
	text = regexp.MustCompile(`\s+`).ReplaceAllString(text, " ")

	// 移除特殊字符
	text = strings.ReplaceAll(text, "&nbsp;", " ")
	text = strings.ReplaceAll(text, "&lt;", "<")
	text = strings.ReplaceAll(text, "&gt;", ">")
	text = strings.ReplaceAll(text, "&amp;", "&")
	text = strings.ReplaceAll(text, "&quot;", "\"")

	return strings.TrimSpace(text)
}

// TruncateText 截取指定长度的文本
func TruncateText(text string, length int) string {
	runeText := []rune(text)
	if len(runeText) <= length {
		return text
	}
	return string(runeText[:length])
}