- tokenization. (These have been begun in
test/LexerTest.php, primarily to ensure delegation to pragma implementations works correctly.)
- Some lingering issues with too much indentation and/or extra newlines; all
- such tests are marked "whitespace-issues"