1. 17
  1.  

  2. [Comment removed by author]

    1. 2

      Are error messages from parser combinator libraries as bad as the ones from yacc?

      Yes. They tend to just give an error for the combinator itself, and provide no larger context.

      1. 2

        Did you perhaps mean to comment on this post instead?

        1. 1

          Yes

      2. 1

        This book is enticing but I’m not sure the appeal is clear to me. The first section seems to provide a quite detailed notation for describing the expected behavior for programs, which would seem to provide a useful lingua franca for precisely discussing specs before writing. But I’ve just read the first section where the author actually puts it all together and it seems as though he expects you to actually write a program in it, and then translate as literally as possible into a programming language in order to ensure correctness. Which seems nutty to me. Heck, in the first and most trivial example it already results in a GOTO, which can’t be a good sign. And I can’t help but imagine that it would be disastrous to take these semantics and then write them in, say, Erlang, thus forgoing all pattern matching—to give one example. I’ll keep reading. I’m interested in other people’s thoughts.