FYI, I tried Zach Elliott's j-mode and it throws a "Variable binding depth exceeds max-specpdl-size" error at startup, which seems to be from the (defconst j-help-dictionary-data-block ...) in j-help.el
The sourceforge version doesn't have those additional components and does not throw that error. -------- That said, the problem with emacs forward-word is apparent at https://www.emacswiki.org/emacs/EmacsSyntaxTable Basically, according to that concept, I think the word forming J characters should be 0..9 A..Z, _, and a..z -- the emacs syntactic simplifications don't work well for J tokens (and : and . should probably be classified as whitespace - they're not, but this probably gets the closest approximation to reasonable behavior that I think you can get from an emacs syntax table). Or, maybe punctuation. Basically, though, it can't be perfect because emacs doesn't understand symbol forming rules (nor word forming rules) where the first character is special. Thanks, -- Raul On Wed, Mar 12, 2014 at 8:35 PM, Joe Bogner <[email protected]> wrote: > On Wed, Mar 12, 2014 at 6:14 PM, Devon McCormick <[email protected]> wrote: >> OK - great! I got it to work though I invoke J through a batch file so I >> can supply the arguments to run it with my usual defaults loaded. > > Excellent > >> >> Are you familiar with Zach Elliott's "j-mode.el" at >> http://github.com/zellio/j-mode ? >> > > Yes, I have used this one more though: > http://j-mode.sourceforge.net/. I can't recall why I settled on that > one over Zach Elliott's > >> Also, are you familiar with this: >> >> ;* disable-substitution-on-recall.el: prevent unwanted interpretation of !: >> in J. >> (defun disable-substitution-on-recall() >> "Avoid unwanted interpretation of '!:' in J." >> (interactive) ; Necessary to be able to run from interactive key >> sequence? >> (setq comint-input-autoexpand nil) >> (setq comint-dynamic-complete-functions nil)) >> ;EG NB. (disable-substitution-on-recall) [C-x C-e] >> >> I think Joey Tuttle showed me this: it prevents emacs from garbling J >> expressions like "6!:2" when they are recalled to a session via >> "comint-previous-input". >> >> > > No, but I don't have any problems recalling input with !: > > For example, > > 6!:2 '1+1' > > I can recall this just fine. Maybe this was an issue in an earlier > version of emacs or J? Or, am I misunderstanding the issue? > > By the way, in the comint-buffer, ctrl-shift up is a nice way to > recall too. It's the same as JHS which is convenient. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
