/**
 * Search-box.ai
 * Author: Simon-Pierre Boucher
 * Contact: contact@spboucher.ai
 * File: packages/shared/src/text.ts
 * Description: Text sanitation for untrusted web content (JSON- and Postgres-safe).
 */

/**
 * Makes arbitrary scraped text safe for JSON serialization and Postgres storage:
 * removes lone UTF-16 surrogates (invalid JSON) and NUL bytes (rejected by PG).
 * Call again after slicing — a cut can split a surrogate pair.
 */
export function sanitizeText(s: string): string {
  return s
    .replace(/[\uD800-\uDBFF](?![\uDC00-\uDFFF])/g, "")
    .replace(/(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]/g, "")
    .replace(/\u0000/g, "");
}