// Link and display policy of the richeditor field (D-10, T-17-10, T-17-11).
// The server sanitizer (seaglass) is the gate on every save; these client
// checks keep the editor from offering what the gate would remove and keep a
// stored value that predates the gate from running in the admin.
import { EDITOR_TAGS, inertFragment } from './lossy'

/** Link schemes the editor applies; anything else must be a relative URL. */
export const LINK_SCHEMES: readonly string[] = ['http', 'https', 'mailto', 'tel']

// Browsers ignore ASCII control characters, spaces and DEL inside a scheme,
// so "java\tscript:" is read with them removed.
const URL_IGNORED = /[\x00-\x20\x7f]+/g
const URL_SCHEME = /^([a-z][a-z0-9+.-]*):/i

/**
 * Whether url may be a link href: an http, https, mailto or tel URL, or a
 * relative one (a path starting with /, #, ? or ./, or a bare relative path).
 */
export function isAllowedLink(url: string | null | undefined): boolean {
  if (url == null) {
    return false
  }
  const cleaned = url.replace(URL_IGNORED, '')
  if (cleaned === '') {
    return false
  }
  const scheme = URL_SCHEME.exec(cleaned)
  if (scheme) {
    return LINK_SCHEMES.includes(scheme[1]!.toLowerCase())
  }
  try {
    new URL(cleaned, 'https://admin.invalid/')
    return true
  } catch {
    return false
  }
}

/** Elements dropped with their content when a value is displayed. */
const DROPPED = new Set([
  'script',
  'style',
  'template',
  'noscript',
  'iframe',
  'frame',
  'frameset',
  'object',
  'embed',
  'applet',
  'svg',
  'math',
  'title',
  'head',
  'meta',
  'link',
  'base',
])

function clean(parent: ParentNode): void {
  for (const node of Array.from(parent.childNodes)) {
    if (node.nodeType === Node.COMMENT_NODE) {
      node.remove()
      continue
    }
    if (node.nodeType !== Node.ELEMENT_NODE) {
      continue
    }
    const element = node as Element
    const name = element.localName.toLowerCase()
    if (DROPPED.has(name)) {
      element.remove()
      continue
    }
    clean(element)
    if (!EDITOR_TAGS.has(name)) {
      // Another element keeps its text: its children take its place.
      element.replaceWith(...Array.from(element.childNodes))
      continue
    }
    for (const attribute of Array.from(element.attributes)) {
      const keep = name === 'a' && attribute.name === 'href' && isAllowedLink(attribute.value)
      if (!keep) {
        element.removeAttribute(attribute.name)
      }
    }
  }
}

/**
 * The read-only rendering of a stored value: only the editor's elements,
 * only allowed link hrefs, no other attribute, no comment. It is what may be
 * bound as HTML in the admin; the stored value itself is never bound.
 */
export function displayHtml(html: string): string {
  if (html === '') {
    return ''
  }
  const fragment = inertFragment(html)
  clean(fragment)
  const holder = document.createElement('div')
  holder.append(fragment)
  return holder.innerHTML
}