News for hackers. Stuff that matters.

jimroskind

Re:The Jim Roskind C/C++ Grammar (2018) on Sunday October 11, 2020 @11:28AM (#24746434)
I doubt there are licensing issues (or at least that is how I TRIED to write the license I included)... but if there are, please contact me on LinkedIn. Are you suggesting I should just put it all on Github as a publicly readable repo? I could probably do so.

Thanks,

Jim

Re:The Jim Roskind C/C++ Grammar (2018) on Friday October 09, 2020 @06:50PM (#24735210)
I'm Jim Roskind, the author of the grammars.

It is an honor to see my grammars still being discussed.

These days I work at Amazon, but I can be found on LinkedIn. If contacted with a comment about the grammars (as opposed to a spam missive from a head hunter), I'll typically respond.

What I really need to (finally) do someday is publish my hacks that I added to byacc to increase usability (I think I called it jryacc).

I've always felt that yacc's "error reports" are kindred to having a compiler error listing limited to:

"you have 7 severe errors, 3 errors, and 23 warnings,"

and a reference:

  "to understand the issues, please read the machine code."
:-)

With yacc, verbose output suggests (at best) that you read the "state transition diagrams" <ugh> to make sense of "n shift reduce, and m reduce-reduce conflicts."

jryacc provided the shortest sequence of terminals to reach any state, identifies any non-terminal that is both left and right recursive (a common malady in grammar construction, which ensures ambiguity), the shortest sequence of terminals to construct each non-terminal, etc.

...oh... and not to fear... I have copies of the grammars and files mentioned... so hopefully they won't be too lost too soon.

Thanks again for the interesting comments.

Jim