On 2018/05/01 20:14:34, thomasmorley651 wrote:
in the light of your findings and David's reply (#17) I changed my
mind about
those "-"-signs and now agree we should not make them a topic here. Anyway, I found the discussion very instructive, helping me to get a
deeper
understanding. I hope it was not too tedious for you.
One really has to understand that the rules in the grammar describe what constructs are accepted and interpreted in what manner. They don't assign a logic or meaning to them. It makes the grammar easier to read when the rules are laid according to a higher-level plan, but the restrictions of an LALR(1) parser generator (which is what the rules are written for) don't leave a lot of leeway of expressing the concepts of LilyPond straightforwardly. In addition, quite a bit of the "inherent logic" has been designed after the fact and then those rules/productions not matching that logic have been meddled with. The grammar is the ultimate source of information about how LilyPond will behave, but not what that means. That's ultimately for the documentation writers to decide. https://codereview.appspot.com/343060043/ _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-devel
