// SEO helpers for meta tags and JSON-LD. Strips HTML and decodes common
// entities so seller-authored rich-text descriptions don't leak markup into
// <meta name="description">, og:description, twitter:description, or
// schema.org Product.description.

const NAMED_ENTITIES: Record<string, string> = {
  nbsp: ' ',
  amp: '&',
  lt: '<',
  gt: '>',
  quot: '"',
  apos: "'",
};

function decodeEntities(text: string): string {
  return text
    .replace(/&(nbsp|amp|lt|gt|quot|apos);/g, (_, name: string) => NAMED_ENTITIES[name] ?? '')
    .replace(/&#(\d+);/g, (_, code: string) => String.fromCodePoint(Number(code)))
    .replace(/&#x([0-9a-f]+);/gi, (_, code: string) => String.fromCodePoint(parseInt(code, 16)));
}

// Strip HTML tags, decode entities, collapse whitespace. Idempotent on plain text.
export function stripHtmlForSeo(input: string | null | undefined): string {
  if (!input) return '';
  return decodeEntities(input.replace(/<[^>]*>/g, ' '))
    .replace(/\s+/g, ' ')
    .trim();
}

// Build a meta description from raw HTML/text.
// Truncates at the nearest word boundary so we never cut mid-word, and
// appends a Unicode ellipsis. Returns '' for null/empty input so callers
// can fall back cleanly.
export function buildMetaDescription(input: string | null | undefined, maxLength = 160): string {
  const stripped = stripHtmlForSeo(input);
  if (!stripped) return '';
  if (stripped.length <= maxLength) return stripped;

  const room = maxLength - 1; // leave room for ellipsis
  const cut = stripped.slice(0, room);
  const lastSpace = cut.lastIndexOf(' ');
  // Only break on word boundary when it's reasonably close to the cap — otherwise
  // a single very long word would shrink the description to almost nothing.
  const safeCut = lastSpace > room * 0.7 ? cut.slice(0, lastSpace) : cut;
  return `${safeCut.trim()}…`;
}