This patch attempts to fix semantic indentation with clisp. I'm not sure how great an idea it is, but it seems to work for me.
It works by changing the macro-functon on common-lisp:defmacro to one that also stores the arglist on the defined macro's symbol-plist. Strictly speaking, I don't believe you're meant to modify symbols in the common-lisp package in this way, but it seems to work with clisp.
Any thoughts would be much appreciated!
2006-06-14 Damyan Pepper damyanp@gmail.com
* swank-clisp.lisp: Replace common-lisp:defmacro with a modified version that stores the argument list in a way that can be retrieved with swank-backend:arglist. This makes semantic indentation work for macros with &body arguments.
$ cvs -q diff -u Index: swank-clisp.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank-clisp.lisp,v retrieving revision 1.58 diff -u -r1.58 swank-clisp.lisp --- swank-clisp.lisp 22 Mar 2006 16:40:01 -0000 1.58 +++ swank-clisp.lisp 14 Jun 2006 11:47:00 -0000 @@ -137,12 +137,13 @@
(defimplementation arglist (fname) (block nil - (or (ignore-errors - (let ((exp (function-lambda-expression fname))) - (and exp (return (second exp))))) - (ignore-errors - (return (ext:arglist fname))) - :not-available))) + (or (get fname 'arglist) + (ignore-errors + (let ((exp (function-lambda-expression fname))) + (and exp (return (second exp))))) + (ignore-errors + (return (ext:arglist fname))) + :not-available)))
(defimplementation macroexpand-all (form) (ext:expand-form form)) @@ -580,6 +581,22 @@ #+lisp=cl (ext:quit) #-lisp=cl (lisp:quit))
+♀ + +;;; Replace cl:defmacro with our own that stores the arglists +(defvar *old-defmacro* (copy-symbol 'common-lisp:defmacro t)) + +(defmacro defmymacro (&environment env &whole whole name params &body body) + (declare (ignore body)) + `(progn + (setf (get ',name 'swank-backend::arglist) ',params) + ,(funcall (macro-function *old-defmacro*) whole env) + ',name)) + +(setf *old-defmacro* (copy-symbol 'common-lisp:defmacro t)) +(setf (macro-function 'common-lisp:defmacro) (macro-function 'defmymacro)) + + ;;; Local Variables: ;;; eval: (put 'compile-file-frobbing-notes 'lisp-indent-function 1) ;;; eval: (put 'dynamic-flet 'common-lisp-indent-function 1)
Damyan Pepper damyanp@gmail.com writes:
+;;; Replace cl:defmacro with our own that stores the arglists +(defvar *old-defmacro* (copy-symbol 'common-lisp:defmacro t))
+(defmacro defmymacro (&environment env &whole whole name params &body body)
- (declare (ignore body))
`(progn(setf (get ',name 'swank-backend::arglist) ',params),(funcall (macro-function *old-defmacro*) whole env)',name))+(setf *old-defmacro* (copy-symbol 'common-lisp:defmacro t)) +(setf (macro-function 'common-lisp:defmacro) (macro-function 'defmymacro))
this is really evil. what are the odds of "fixing" clisp instead?
Marco Baringer mb@bese.it writes:
Damyan Pepper damyanp@gmail.com writes:
+;;; Replace cl:defmacro with our own that stores the arglists
this is really evil. what are the odds of "fixing" clisp instead?
It's interesting to know that it's as evil as I thought it probably was!
I've spent a bit of time trying to find the appropriate place in the clisp source to do this, but it's a very alien place to me. I'll have an ask on the clisp lists to see what the thoughts there are.
Thanks,
Damyan.
Damyan Pepper damyanp@gmail.com writes:
This patch attempts to fix semantic indentation with clisp. I'm not sure how great an idea it is, but it seems to work for me.
It works by changing the macro-functon on common-lisp:defmacro to one that also stores the arglist on the defined macro's symbol-plist. Strictly speaking, I don't believe you're meant to modify symbols in the common-lisp package in this way, but it seems to work with clisp.
Any thoughts would be much appreciated!
I don't use Clisp, so I'm hardly qualified to comment, but have you considered offering a similar patch against Clisp itself?
Cheers,
-- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs."
Nikodemus Siivola nikodemus@random-state.net writes:
Damyan Pepper damyanp@gmail.com writes:
I don't use Clisp, so I'm hardly qualified to comment, but have you considered offering a similar patch against Clisp itself?
Yup, I'm having a look into doing that now.
-
Damyan Pepper -
Marco Baringer -
Nikodemus Siivola