Fixes more cases, found by @stigrj, where the grammar was not giving the expected results. See #35 and #30. It also fixes the missing files from the installed package.
I have refactored the recursive grammar a bit, also to avoid code duplication. I have no love lost for pyparsing, which is a very strange beast indeed. Ordering of tokens is important both for matching and performance. With the current refactoring tests are noticeably faster.
Fixes more cases, found by @stigrj, where the grammar was not giving the expected results. See #35 and #30. It also fixes the missing files from the installed package. I have refactored the recursive grammar a bit, also to avoid code duplication. I have no love lost for pyparsing, which is a very strange beast indeed. Ordering of tokens is important both for matching and performance. With the current refactoring tests are noticeably faster.