/*******
 * Normalize HTML void elements to an explicit self-closing form (`/>`).
 *
 * This rewrites opening tags for known “void” HTML elements (elements that do
 * not have closing tags) into a consistent self-closing representation:
 *
 * - `<img src="x">`   → `<img src="x" />`
 * - `<br>`            → `<br />`
 *
 * What it targets:
 * - Only tag names in `voidTags` (case-insensitive match).
 * - Only tags that are not already self-closed (i.e. not `<img .../>`).
 *
 * What it avoids:
 * - It does not touch non-void tags.
 * - It attempts to avoid rewriting tags that already have an explicit closer
 *   immediately following (e.g. `<tag ...></tag>`), though for true void tags
 *   that pattern is already “nonstandard HTML”.
 *
 * Rationale:
 * - Some downstream tokenizers/serializers (including Hson-ish pipelines) find
 *   it simpler if void elements always appear in a single, unambiguous form.
 *
 * Caveats:
 * - This is a regex-based normalizer, not a full HTML parser.
 * - The `voidTags` list is the authority; tags not listed will never be
 *   rewritten, even if they are void in some HTML dialect.
 *
 * @param $input - HTML source text to normalize.
 * @returns HTML with matching void tags rewritten as `<tag ... />`.
 *******/
export declare function expand_void_tags(src: string): string;
//# sourceMappingURL=expand-self-closing.d.ts.map