@spec extract_html_content(String.t(), non_neg_integer()) ::
  {String.t(), non_neg_integer(), non_neg_integer()}

Extract HTML content starting from position until we hit a JSON structural delimiter. Returns {html_content, chars_consumed, bytes_consumed}.

Strategy: Track HTML tag depth. Only stop at JSON delimiters when:

• We're at HTML tag depth 0 (all tags closed)
• We're at JSON depth 0 (no nested JSON-like braces)
• We're not inside an HTML tag marker (between < and >)

@spec is_html_start?(String.t(), non_neg_integer()) :: boolean()

Check if the current position starts with HTML content. Returns true if it looks like HTML (starts with < followed by tag name or !DOCTYPE).

@spec process_html_iolist(String.t(), map()) ::
  {iodata(), non_neg_integer(), non_neg_integer(), list()}

Process HTML content for IO list version. Returns {html_iolist, chars_consumed, bytes_consumed, repairs}.

@spec process_html_string(String.t(), map()) ::
  {String.t(), non_neg_integer(), non_neg_integer(), list()}

Process HTML content for regular string version. Returns {html_string, chars_consumed, bytes_consumed, repairs}.

@spec quote_html_content(String.t(), non_neg_integer()) :: {String.t(), list()}

Process HTML content by wrapping it in quotes and escaping internal quotes and special characters. Returns {quoted_html, repairs}.