Improve CddlLexer performance

[blu-base]

The CddlLexer always seemed incredibly slow for the tiny provided example file. Profiling showed that the re engine took most of the time. After some testing the _re_id expression proved to be the culprit.

An other minor change improved grouping of the String.Single Token. That's the only reason the examplefile output needs an update.

---

If you want, I can remove the modification for the String.Single token to only keep the important change to the _re_id expression

[blu-base]

The failed check seems to be a fluke?

  =================================== FAILURES ===================================
  ________________________ test_happy_javascript_fragment ________________________
  
  lexer_html = <pygments.lexers.HtmlLexer>
  
      def test_happy_javascript_fragment(lexer_html):
          """valid, even long Javascript fragments should still get parsed ok"""
      
          fragment = "<script type=\"text/javascript\">"+"alert(\"hi\");"*2000+"</script>"
          start_time = time.time()
          tokens = list(lexer_html.get_tokens(fragment))
          assert all(x[1] != Token.Error for x in tokens)
  >       assert time.time() - start_time < MAX_HL_TIME, \
              'The HTML lexer might have an expensive happy-path script case'
  E       AssertionError: The HTML lexer might have an expensive happy-path script case
  E       assert (1637013420.331315 - 1637013409.7813637) < 10
  E        +  where 1637013420.331315 = <built-in function time>()
  E        +    where <built-in function time> = time.time
  
  /home/runner/work/pygments/pygments/tests/test_html_lexer.py:31: AssertionError
  =========================== short test summary info ============================
  FAILED tests/test_html_lexer.py::test_happy_javascript_fragment - AssertionEr...
  ============ 1 failed, 3775 passed, 12 skipped in 294.47s (0:04:54) ============

[birkenfeld]

Looks good to me (and I like the cleanup for the string case), thanks!