Archipelago — open-source initial import

aiui/packages/app/src/utils/html.ts

@@ -0,0 +1,40 @@
+/** Check if a URL has a safe http(s) protocol */
+export function isSafeImgSrc(src: string): boolean {
+  try {
+    const u = new URL(src)
+    return /^https?:$/i.test(u.protocol)
+  } catch {
+    return false
+  }
+}
+
+/** Type-guard variant: check if value is a defined, safe http(s) URL */
+export function isSafeUrl(u: string | undefined): u is string {
+  return !!u && typeof u === 'string' && /^https?:\/\//i.test(u.trim())
+}
+
+import DOMPurify from 'dompurify'
+
+const ALLOWED_TAGS_LIST = ['p', 'br', 'a', 'strong', 'em', 'b', 'i', 'ul', 'ol', 'li', 'blockquote', 'h1', 'h2', 'h3', 'h4', 'span', 'div']
+
+/** Sanitize HTML using DOMPurify with restricted tag set */
+export function sanitizeHtml(html: string): string {
+  return DOMPurify.sanitize(html, {
+    ALLOWED_TAGS: ALLOWED_TAGS_LIST,
+    ALLOWED_ATTR: ['href', 'src', 'target', 'rel'],
+  })
+}
+
+/** Escape HTML entities for safe rendering in a text context */
+export function escapeHtml(text: string): string {
+  return text.replace(/&/g, '&amp;').replace(/</g, '&lt;').replace(/>/g, '&gt;')
+}
+
+/** Extract a domain from a URL, stripping www prefix */
+export function formatDomain(url: string): string {
+  try {
+    return new URL(url).hostname.replace(/^www\./, '')
+  } catch {
+    return url
+  }
+}