class Regexp::Lexer

def break_codepoint_list(token)

c.f. #break_literal.
to the last codepoint, e.g. /\u{61 62 63}{3}/ =~ 'abccc'
if a codepoint list is followed by a quantifier, that quantifier applies 
def break_codepoint_list(token)
  lead, _, tail = token.text.rpartition(' ')
  return if lead.empty?
  token_1 = Regexp::Token.new(:escape, :codepoint_list, lead + '}',
            token.ts, (token.te - tail.length),
            nesting, set_nesting, conditional_nesting)
  token_2 = Regexp::Token.new(:escape, :codepoint_list, '\u{' + tail,
            (token.ts + lead.length + 1), (token.te + 3),
            nesting, set_nesting, conditional_nesting)
  self.shift = shift + 3 # one space less, but extra \, u, {, and }
  token_1.previous = preprev_token
  token_1.next = token_2
  token_2.previous = token_1 # .next will be set by #lex
  [token_1, token_2]
end