October 2004 - slime-devel - mailman3.common-lisp.net

[slime-devel] cl-indent lossage
by Lynn Quam 29 Oct '04

29 Oct '04

Was it intentional that the version of cl-indent.el in SLIME changed the indentation for the LOOP macro? In particular cl-indent in SLIME now does this indentation: (loop for x in l for y for = (foo x) do (mumble x y)) whereas cl-indent in EMACS does this indentation: (loop for x in l for y for = (foo x) do (mumble x y)) I much prefer the EMACS indentation.

2 4

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 29 Oct '04

29 Oct '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.560 slime/ChangeLog:1.562 --- slime/ChangeLog:1.560 Thu Oct 28 23:41:38 2004 +++ slime/ChangeLog Fri Oct 29 00:19:03 2004 @@ -1,5 +1,7 @@ 2004-10-28 Helmut Eller <heller(a)common-lisp.net> + * swank-clisp.lisp: Ups. Undo previous change. + * swank-clisp.lisp: Add workaround for CLISP's broken control string parser. @@ -39,6 +41,8 @@ [standard-method]: Use less than 80 columns. (inspector-call-nth-action): Don't accept &rest args. Was never used. + (inspect-for-emacs) [integer]: Fix control string. Thanks to CSR + for pointing it out. 2004-10-27 Helmut Eller <heller(a)common-lisp.net>

1 0

[slime-devel] "Unprofile All" does not work in CMUCL 19a
by Chisheng Huang 29 Oct '04

29 Oct '04

Slime > Profiling > Unprofile All does not set profile:*timed-functions* to NIL. I'm using CMUCL 19a Linux/x86, GNU Emacs 21.3.1, and Slime cvs-updated at 2004-10-25 15:38Z. Thanks a lot. -cph

3 2

Re: [slime-devel] CLISP support broken in CVS
by Bill_Clementson＠peoplesoft.com 29 Oct '04

29 Oct '04

Attached is a patch that Helmut Eller made to CLISP to correct format string functionality. Thanks Helmut for the fix! - Bill ----- Forwarded by Bill Clementson/PeopleSoft on 10/28/2004 06:07 PM ----- "Helmut Eller" <e9626484(a)stud3.tuwien.a To: Bill_Clementson(a)peoplesoft.com c.at> cc: slime-devel(a)common-lisp.net Sent by: Subject: Re: [slime-devel] CLISP support broken in CVS slime-devel-bounces@comm on-lisp.net 10/28/2004 03:46 PM Bill_Clementson(a)peoplesoft.com writes: > FYI - CLISP 2.33 support seems to be broken in CVS. Here's the contents of > the *inferior-lisp* buffer when I attempt to load CLISP: [snip] > Current point in control string: > Value: ~D = #x~X = #o~O = #b~:,,' ,8B = ~E > | > Break 1 SWANK[2]> I think this control string is legal and CLISP is broken. I added a IMO fixed version of CLISP's format string parser to swank-clisp.lisp. The patch against the CLISP source file is attached below, in case somebody has the motivation to report the bug to the CLISP developers. Helmut (See attached file: format.lisp.patch) _______________________________________________ slime-devel site list slime-devel(a)common-lisp.net http://common-lisp.net/mailman/listinfo/slime-devel

1 0

[slime-devel] CLISP support broken in CVS
by Bill_Clementson＠peoplesoft.com 28 Oct '04

28 Oct '04

FYI - CLISP 2.33 support seems to be broken in CVS. Here's the contents of the *inferior-lisp* buffer when I attempt to load CLISP: [1]> ;; Loading file C:\usr\home\site\slime\swank-loader.lisp ... ;; Compiling C:\usr\home\site\slime\swank-backend.lisp... Compiling file C:\usr\home\site\slime\swank-backend.lisp ... WARNING in DEFINTERFACE-GEN-DEFAULT-IMPL in lines 98..119 : variable RECEIVED-ARGS is not used. Misspelled or missing IGNORE declaration? WARNING in (DEFINTERFACE EMACS-CONNECTED (STREAM) ...)-22-3-1-1 in lines 208..218 : variable STREAM is not used. Misspelled or missing IGNORE declaration? WARNING in (DEFINTERFACE MAKE-STREAM-INTERACTIVE (STREAM) ...)-40-3-1-1 in lines 342..348 : variable STREAM is not used. Misspelled or missing IGNORE declaration? WARNING in (DEFINTERFACE FRAME-PACKAGE (FRAME) ...)-50-3-1-1 in lines 443..446 : variable FRAME is not used. Misspelled or missing IGNORE declaration? WARNING in (DEFINTERFACE CONDITION-REFERENCES (CONDITION) ...)-60-3-1-1 in lines 502..509 : variable CONDITION is not used. Misspelled or missing IGNORE declaration? WARNING in (DEFINTERFACE CONDITION-EXTRAS (CONDITION) ...)-61-3-1-1 in lines 511..515 : variable CONDITION is not used. Misspelled or missing IGNORE declaration? WARNING in (DEFINTERFACE BUFFER-FIRST-CHANGE (FILENAME) ...)-71-3-1-1 in lines 556..558 : variable FILENAME is not used. Misspelled or missing IGNORE declaration? Wrote file C:\usr\home\.slime\fasl\clisp-2.33\swank-backend.fas ;; Loading file C:\usr\home\.slime\fasl\clisp-2.33\swank-backend.fas ... ;; Loaded file C:\usr\home\.slime\fasl\clisp-2.33\swank-backend.fas ;; Compiling C:\usr\home\site\slime\nregex.lisp... Compiling file C:\usr\home\site\slime\nregex.lisp ... Wrote file C:\usr\home\.slime\fasl\clisp-2.33\nregex.fas ;; Loading file C:\usr\home\.slime\fasl\clisp-2.33\nregex.fas ... ;; Loaded file C:\usr\home\.slime\fasl\clisp-2.33\nregex.fas ;; Compiling C:\usr\home\site\slime\xref.lisp... Compiling file C:\usr\home\site\slime\xref.lisp ... Wrote file C:\usr\home\.slime\fasl\clisp-2.33\xref.fas ;; Loading file C:\usr\home\.slime\fasl\clisp-2.33\xref.fas ... ;; Loaded file C:\usr\home\.slime\fasl\clisp-2.33\xref.fas ;; Compiling C:\usr\home\site\slime\metering.lisp... Compiling file C:\usr\home\site\slime\metering.lisp ... Wrote file C:\usr\home\.slime\fasl\clisp-2.33\metering.fas ;; Loading file C:\usr\home\.slime\fasl\clisp-2.33\metering.fas ... ;; Loaded file C:\usr\home\.slime\fasl\clisp-2.33\metering.fas ;; Compiling C:\usr\home\site\slime\swank-clisp.lisp... Compiling file C:\usr\home\site\slime\swank-clisp.lisp ... WARNING: DEFMETHOD: redefining method (LISP-IMPLEMENTATION-TYPE-NAME NIL) in C:\usr\home\site\slime\swank-clisp.lisp, was defined in top-level WARNING: DEFMETHOD: redefining method (UNPROFILE-ALL NIL) in C:\usr\home\site\slime\swank-clisp.lisp, was defined in top-level WARNING in COMPILER-NOTE-LOCATION in lines 374..386 : *BUFFER-NAME* is neither declared nor bound, it will be treated as if it were declared SPECIAL. WARNING in COMPILER-NOTE-LOCATION in lines 374..386 : *BUFFER-NAME* is neither declared nor bound, it will be treated as if it were declared SPECIAL. WARNING in COMPILER-NOTE-LOCATION in lines 374..386 : *BUFFER-OFFSET* is neither declared nor bound, it will be treated as if it were declared SPECIAL. Wrote file C:\usr\home\.slime\fasl\clisp-2.33\swank-clisp.fas ;; Loading file C:\usr\home\.slime\fasl\clisp-2.33\swank-clisp.fas ... WARNING: Replacing method #<STANDARD-METHOD (#<BUILT-IN-CLASS T>)> in #<GENERIC-FUNCTION CALL-WITHOUT-INTERRUPTS> WARNING: Replacing method #<STANDARD-METHOD NIL> in #<GENERIC-FUNCTION LISP-IMPLEMENTATION-TYPE-NAME> WARNING: Replacing method #<STANDARD-METHOD (#<BUILT-IN-CLASS T>)> in #<GENERIC-FUNCTION SET-DEFAULT-DIRECTORY> WARNING: Replacing method #<STANDARD-METHOD (#<BUILT-IN-CLASS T>)> in #<GENERIC-FUNCTION ARGLIST> WARNING: Replacing method #<STANDARD-METHOD NIL> in #<GENERIC-FUNCTION UNPROFILE-ALL> WARNING: DEFUN/DEFMACRO: redefining function BREAK in C:\usr\home\.slime\fasl\clisp-2.33\swank-clisp.fas, was defined in D:\gnu\clisp\sf\clisp\build-O-mingw\condition.fas ;; Loaded file C:\usr\home\.slime\fasl\clisp-2.33\swank-clisp.fas ;; Compiling C:\usr\home\site\slime\swank-gray.lisp... Compiling file C:\usr\home\site\slime\swank-gray.lisp ... Wrote file C:\usr\home\.slime\fasl\clisp-2.33\swank-gray.fas ;; Loading file C:\usr\home\.slime\fasl\clisp-2.33\swank-gray.fas ... ;; Loaded file C:\usr\home\.slime\fasl\clisp-2.33\swank-gray.fas ;; Compiling C:\usr\home\site\slime\swank.lisp... Compiling file C:\usr\home\site\slime\swank.lisp ... *** - Non-existent directive Current point in control string: Value: ~D = #x~X = #o~O = #b~:,,' ,8B = ~E | Break 1 SWANK[2]>

3 3

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 28 Oct '04

28 Oct '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.559 slime/ChangeLog:1.560 --- slime/ChangeLog:1.559 Wed Oct 27 12:59:06 2004 +++ slime/ChangeLog Thu Oct 28 23:41:38 2004 @@ -1,3 +1,45 @@ +2004-10-28 Helmut Eller <heller(a)common-lisp.net> + + * swank-clisp.lisp: Add workaround for CLISP's broken control + string parser. + + * swank-cmucl.lisp (set-step-breakpoints): Handle breakpoints at + single-return points in escaped frames better. Previously we + tried to set a breakpoint at the current position and consequently + was only hit during the next call. + (inspect-for-emacs)[function]: Call the next method only for + funcallable instances. + (profile-report, profile-reset, unprofile-all): We have to use + eval because the macro expansion depends on the value of + *timed-functions*. Reported by Chisheng Huang. + + * slime.el (slime-space): Call slime-message in the right buffer, + so that after-command hooks are added in the right buffer. + Reported by Juho Snellman. + (slime-dispatch-event): Accept stepping flag. + (sldb-setup): Don't query when entering a recursive edit. + (sldb-exit): Don't kill the buffer if we are in stepping mode. + (slime-inspector-insert-ispec): New function. + (slime-open-inspector): Use it. + (slime-inspector-operate-on-point): Simplified. + (test interactive-eval): Fix test case. + (slime-kill-all-buffers): More regexp kludges. From Bill Clementson. + + * swank-backend.lisp (activate-stepping): New function. + + * swank.lisp (*sldb-stepping-p*): New variable. Used to tell + emacs that the debugger buffer should not be closed even if we + unwind. + (debug-in-emacs): Use it. + (sldb-step): Moved to the front end. + (inspector-princ, method-specializers-for-inspect): Simplified. + (methods-by-applicability): Use a simpler algorithm. I doubt there + is much difference in practice. + (inspect-for-emacs)[symbol, function, standard-generic-function] + [standard-method]: Use less than 80 columns. + (inspector-call-nth-action): Don't accept &rest args. Was never + used. + 2004-10-27 Helmut Eller <heller(a)common-lisp.net> * swank-sbcl.lisp (signal-compiler-condition): Actually delete one

1 0

Re: [slime-devel] Sayoonara command doesn't close repl buffer
by Bill_Clementson＠peoplesoft.com 28 Oct '04

28 Oct '04

This patch I sent doesn't appear to have been put into CVS. - Bill Bill_Clementson@peopleso ft.com To: slime-devel(a)common-lisp.net Sent by: cc: (bcc: Bill Clementson/PeopleSoft) slime-devel-bounces@comm Subject: [slime-devel] Sayoonara command doesn't close repl buffer on-lisp.net 10/26/2004 07:04 PM I just upgraded from a late-September SLIME (yes, I know that's pretty slack :) ) to CVS and noticed something different. The Sayoonara REPL command (when you type ",s<enter>" in the repl buffer) closes the CL implementation but not the slime repl buffer. I did notice that the repl buffer now no longer has a numeric sufffix - instead it's If it's suffixed with the CL implementatin name. Since I haven't been following the mailing list very closely lately, I'm not sure whether this behaviour is intentional or not. If it is not intentional, the following patch should fix the problem: *** c:/DOCUME~1/BC1886~1/LOCALS~1/Temp/slime-orig.el Tue Oct 26 19:02:52 2004 --- c:/DOCUME~1/BC1886~1/LOCALS~1/Temp/slime.el Tue Oct 26 19:02:52 2004 *************** *** 3108,3113 **** --- 3108,3114 ---- (when (or (string= (buffer-name buf) slime-event-buffer-name) (string-match "^\\*inferior-lisp*" (buffer-name buf)) (string-match "^\\*slime-repl\\[[0-9]+\\]\\*$" (buffer-name buf)) + (string-match (concat "^\\"(format "*slime-repl %s*" (slime-connection-name))) (buffer-name buf)) (string-match "^\\*sldb .*\\*$" (buffer-name buf))) (kill-buffer buf)))) -- Bill Clementson _______________________________________________ slime-devel site list slime-devel(a)common-lisp.net http://common-lisp.net/mailman/listinfo/slime-devel

2 1

[slime-devel] multi-line slime-space windows not closed
by Juho Snellman 28 Oct '04

28 Oct '04

The arglist window that slime-space displays isn't automatically closed on Xemacs 21.4 and CVS slime (2004-10-27). The problem seems to be that slime-message is called from a slime-eval-async callback, where the current buffer is " *cl-connection*". This means that the window-closing lambda is pushed into the buffer-local slime-pre-command-actions of the wrong buffer, and thus never gets called. -- Juho Snellman

2 1

[slime-devel] openmcl defclass documentation workaroud
by Marco Baringer 28 Oct '04

28 Oct '04

the following code, put in openmcl-init.lisp, fixes a bug in openmcl's defclass macro and should allow the inspector to work properly. it's just a dirty hack till the next official release of openmcl. ;;;; defclass work around (defclass common-lisp-user::!@$% () () (:documentation "x")) (when (consp (documentation (find-class 'common-lisp-user::!@$%) t)) (handler-bind ((error (lambda (c) (declare (ignore c)) (invoke-restart 'continue)))) (defmacro defclass (class-name superclasses slots &rest class-options &environment env) (flet ((duplicate-options (where) (signal-program-error "Duplicate options in ~S" where)) (illegal-option (option) (signal-program-error "Illegal option ~s" option)) (make-initfunction (form) (cond ((or (eq form 't) (equal form ''t)) '(function true)) ((or (eq form 'nil) (equal form ''nil)) '(function false)) (t `(function (lambda () ,form)))))) (setq class-name (require-type class-name '(and symbol (not null)))) (setq superclasses (mapcar #'(lambda (s) (require-type s 'symbol)) superclasses)) (let* ((options-seen ()) (signatures ()) (slot-names)) (flet ((canonicalize-defclass-option (option) (let* ((option-name (car option))) (if (member option-name options-seen :test #'eq) (duplicate-options class-options) (push option-name options-seen)) (case option-name (:default-initargs (let ((canonical ())) (let (key val (tail (cdr option))) (loop (when (null tail) (return nil)) (setq key (pop tail) val (pop tail)) (push ``(,',key ,',val ,,(make-initfunction val)) canonical)) `(':direct-default-initargs (list ,@(nreverse canonical)))))) (:metaclass (unless (and (cadr option) (typep (cadr option) 'symbol)) (illegal-option option)) `(:metaclass ',(cadr option))) (:documentation (let* ((doc (cadr option))) (if (and doc (not (typep doc 'string))) (illegal-option option)) `(:documentation ',doc))) (t (list `',option-name `',(cdr option)))))) (canonicalize-slot-spec (slot) (if (null slot) (signal-program-error "Illegal slot NIL")) (if (not (listp slot)) (setq slot (list slot))) (let* ((slot-name (require-type (car slot) 'symbol)) (initargs nil) (other-options ()) (initform nil) (initform-p nil) (initfunction nil) (type nil) (type-p nil) (allocation nil) (allocation-p nil) (documentation nil) (documentation-p nil) (readers nil) (writers nil)) (when (memq slot-name slot-names) (SIGNAL-PROGRAM-error "Duplicate slot name ~S" slot-name)) (push slot-name slot-names) (do ((options (cdr slot) (cddr options)) name) ((null options)) (when (null (cdr options)) (signal-program-error "Illegal slot spec ~S" slot)) (case (car options) (:reader (setq name (cadr options)) (push name signatures) (push name readers)) (:writer (setq name (cadr options)) (push name signatures) (push name writers)) (:accessor (setq name (cadr options)) (push name signatures) (push name readers) (push `(setf ,name) signatures) (push `(setf ,name) writers)) (:initarg (push (require-type (cadr options) 'symbol) initargs)) (:type (if type-p (duplicate-options slot) (setq type-p t)) ;(when (null (cadr options)) (signal-program-error "Illegal options ~S" options)) (setq type (cadr options))) (:initform (if initform-p (duplicate-options slot) (setq initform-p t)) (let ((option (cadr options))) (setq initform `',option initfunction (if (constantp option) `(constantly ,option) `#'(lambda () ,option))))) (:allocation (if allocation-p (duplicate-options slot) (setq allocation-p t)) (setq allocation (cadr options))) (:documentation (if documentation-p (duplicate-options slot) (setq documentation-p t)) (setq documentation (cadr options))) (t (let* ((pair (or (assq (car options) other-options) (car (push (list (car options)) other-options))))) (push (cadr options) (cdr pair)))))) `(list :name ',slot-name ,@(when allocation `(:allocation ',allocation)) ,@(when initform-p `(:initform ,initform :initfunction ,initfunction)) ,@(when initargs `(:initargs ',initargs)) ,@(when readers `(:readers ',readers)) ,@(when writers `(:writers ',writers)) ,@(when type-p `(:type ',type)) ,@(when documentation-p `(:documentation ,documentation)) ,@(mapcan #'(lambda (opt) `(',(car opt) ',(cdr opt))) other-options))))) (let* ((direct-superclasses (or superclasses '(standard-object))) (direct-slot-specs (mapcar #'canonicalize-slot-spec slots)) (other-options (apply #'append (mapcar #'canonicalize-defclass-option class-options )))) `(progn (eval-when (:compile-toplevel) (%compile-time-defclass ',class-name ,env) (progn ,@(mapcar #'(lambda (s) `(note-function-info ',s nil ,env)) signatures))) (ensure-class-for-defclass ',class-name :direct-superclasses ',direct-superclasses :direct-slots ,`(list ,@direct-slot-specs) ,@other-options))))))))) hth. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen

1 0

[slime-devel] SLIME + ACL7
by Julian Stecklina 27 Oct '04

27 Oct '04

Hello, with ACL 7.0 out in the open now I would like to ask, if anyone can comment on how well it works with slime. Can Slime still be configured to act as (X)Eli replacement? If so, does the procedure differ from 6.2? Regards, -- ____________________________ Julian Stecklina / _________________________/ ________________/ / \_________________/ LISP - truly beautiful

1 0