Robert's comments inspire me to mention a very attractive aspect of
Haskell, where documents can be written such that everything *not*
marked as a line of code is treated as a comment. The result is a
"literate" style that is both very easy on the eye and valid input to
a compiler.

It seems easy enough to write a preprocessor that will allow J
definitions to be identified and executed within text that is
otherwise ignored. As for how J sentences would be best identified,
and continuation handled where wrap-around occurs, debate within the J
community would be the next step. (Perhaps double leading
right-paren?)

Tracy


On Tue, Feb 24, 2009 at 1:45 AM, Robert Raschke <[email protected]> wrote:
>
> This reminds me a lot of the pretty-printed Literate Programming
> approach (CWEB, for example). In this style of LP programming (there's
> also the non-pretty-print variety, e.g., noweb), there is a clear
> division between writing the code and its subsequent presentation. And
> since the focus of LP style programming is on exposition of thoughts,
> rather than code-with-comments, this might be an easy way into
> experimenting with various ways of presenting J to a reader.
>
> Robby
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to