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

[slime-devel] Re: Automatic "semantic" indentation
by Bill_Clementson＠peoplesoft.com 05 Apr '04

05 Apr '04

I downloaded the latest SLIME with the new macro indentation code from CVS. It seems to work under LW & CLISP on Windows; however, I get the following error under ACL on Windows: Received signal number 11 (Segmentation violation) [Condition of type SYNCHRONOUS-OPERATING-SYSTEM-SIGNAL] Restarts: 0: [ABORT] Return to sldb level 1. 1: [ABORT] Abort handling SLIME request. 2: [ABORT] Abort entirely from this process. Backtrace: 0: (SWANK::DEBUG-IN-EMACS #<SYNCHRONOUS-OPERATING-SYSTEM-SIGNAL @ #x2064080a>) 1: ((FLET SWANK::SWANK-DEBUGGER-HOOK SWANK::DEBUG-IT)) 2: (SWANK::SWANK-DEBUGGER-HOOK #<SYNCHRONOUS-OPERATING-SYSTEM-SIGNAL @ #x2064080a> #<Function SWANK-DEBUGGER-HOOK>) 3: (EXCL::INTERNAL-INVOKE-DEBUGGER "Error" #<SYNCHRONOUS-OPERATING-SYSTEM-SIGNAL @ #x2064080a> T) 4: (ERROR SYNCHRONOUS-OPERATING-SYSTEM-SIGNAL :NAME #1="Segmentation violation" :NUMBER 11 :FORMAT-CONTROL "Received signal number ~s ~@[(~a)~]" :FORMAT-ARGUMENTS (11 #1#)) 5: (EXCL::HANDLE-PENDING-SIGNAL 11) 6: (SYSTEM::..CONTEXT-SAVING-RUNTIME-OPERATION) 7: (EXCL::LIST-REMOVE-EQL &WHOLE ((COMPILER::VAR COMPILER::NEXT LIST) . COMPILER::BODY)) 8: (SWANK::MACRO-INDENTATION ((COMPILER::VAR COMPILER::NEXT LIST) . COMPILER::BODY)) 9: (SWANK::SYMBOL-INDENTATION COMPILER::DO-QUAD-LIST-TO-END) 10: ((FLET SWANK::UPDATE-INDENTATION/DELTA-FOR-EMACS SWANK::CONSIDER) COMPILER::DO-QUAD-LIST-TO-END) 11: ((FLET SWANK::UPDATE-INDENTATION/DELTA-FOR-EMACS #:G1957) COMPILER::DO-QUAD-LIST-TO-END COMPILER::DO-QUAD-LIST-TO-END) 12: (MAPHASH #<Closure (FLET SWANK::UPDATE-INDENTATION/DELTA-FOR-EMACS #:G1957) @ #x9b2e71a> #<SYMBOL hash-table (sans values) with 2951 entries @ #x200e94ca>) 13: (SWANK::UPDATE-INDENTATION/DELTA-FOR-EMACS #<EQ hash-table with 30 entries @ #x205a3c1a> (#<The ACL-SOCKET package> #<The ACLMOP package> #<The COMMON-LISP package> #<The COMMON-LISP-USER package> #<The COMPILER package> #<The CROSS-REFERENCE package> #<The DEBUGGER package> #<The DEFSYSTEM package> #<The EXCL package> #<The EXCL.SCM package> #<The FOREIGN-FUNCTIONS package> #<The G10 package> #<The GARBAGE package> #<The INSPECT package> #<The KEYWORD package> #<The LEP package> #<The LEP-IO package> #<The MULTIPROCESSING package> #<The NET.URI package> #<The PROFILER package> #<The SWANK package> #<The SWANK-BACKEND package> #<The SWANK-COMPLETION-TEST package> #<The SWANK-IO-PACKAGE package> #<The SWANK-LOADER package> #<The SYSTEM package> #<The TOP-LEVEL package>)) 14: (SWANK::PERFORM-INDENTATION-UPDATE #S(SWANK::CONNECTION :SOCKET-IO #<MULTIVALENT #2=stream socket #3=connected from #1=localhost/4005 to #1#/1722 @ #x205b9392> :DEDICATED-OUTPUT #4=#<MULTIVALENT #2# #3#connected from #1#/1723 to #1#/1724 @ #x205a41aa> :USER-INPUT #<SWANK-BACKEND::SLIME-INPUT-STREAM @ #x205a4a3a> :USER-OUTPUT #4# :USER-IO #<TWO-WAY-STREAM @ #x205a4a22> :CONTROL-THREAD #<MULTIPROCESSING:PROCESS control-thread(5f8)[4] receive @ #x205b824a> :READER-THREAD #<MULTIPROCESSING:PROCESS reader-thread(72c)[5] waiting for input @ #x205b81c2> :READ #<Function READ-FROM-CONTROL-THREAD> :SEND #<Function SEND-TO-CONTROL-THREAD> :SERVE-REQUESTS #<Function SPAWN-THREADS-FOR-CONNECTION> :CLEANUP NIL :INDENTATION-CACHE #<EQ hash-table with 30 entries @ #x205a3c1a> :INDENTATION-CACHE-PACKAGES NIL) (#<The ACL-SOCKET package> #<The ACLMOP package> #<The COMMON-LISP package> #<The COMMON-LISP-USER package> #<The COMPILER package> #<The CROSS-REFERENCE package> #<The DEBUGGER package> #<The DEFSYSTEM package> #<The EXCL package> #<The EXCL.SCM package> #<The FOREIGN-FUNCTIONS package> #<The G10 package> #<The GARBAGE package> #<The INSPECT package> #<The KEYWORD package> #<The LEP package> #<The LEP-IO package> #<The MULTIPROCESSING package> #<The NET.URI package> #<The PROFILER package> #<The SWANK package> #<The SWANK-BACKEND package> #<The SWANK-COMPLETION-TEST package> #<The SWANK-IO-PACKAGE package> #<The SWANK-LOADER package> #<The SYSTEM package> #<The TOP-LEVEL package>)) 15: (SWANK::UPDATE-CONNECTION-INDENTATION #S(SWANK::CONNECTION :SOCKET-IO #<MULTIVALENT #2=stream socket #3=connected from #1=localhost/4005 to #1#/1722 @ #x205b9392> :DEDICATED-OUTPUT #4=#<MULTIVALENT #2# #3#connected from #1#/1723 to #1#/1724 @ #x205a41aa> :USER-INPUT #<SWANK-BACKEND::SLIME-INPUT-STREAM @ #x205a4a3a> :USER-OUTPUT #4# :USER-IO #<TWO-WAY-STREAM @ #x205a4a22> :CONTROL-THREAD #<MULTIPROCESSING:PROCESS control-thread(5f8)[4] receive @ #x205b824a> :READER-THREAD #<MULTIPROCESSING:PROCESS reader-thread(72c)[5] waiting for input @ #x205b81c2> :READ #<Function READ-FROM-CONTROL-THREAD> :SEND #<Function SEND-TO-CONTROL-THREAD> :SERVE-REQUESTS #<Function SPAWN-THREADS-FOR-CONNECTION> :CLEANUP NIL :INDENTATION-CACHE #<EQ hash-table with 30 entries @ #x205a3c1a> :INDENTATION-CACHE-PACKAGES NIL)) 16: (SWANK::SYNC-STATE-TO-EMACS) 17: (SWANK::EVAL-FOR-EMACS (SWANK:BACKTRACE 2 41) NIL 4) 18: (SYSTEM::..RUNTIME-OPERATION . :UNKNOWN-ARGS) 19: (FUNCALL SWANK::EVAL-FOR-EMACS (SWANK:BACKTRACE 2 41) NIL 4) --more-- Don't know whether this is an ACL/Windows-specific error or an ACL-specific error. Does anyone else get this error with the new macro-indentation code under ACL (Windows or Linux)? -- Bill Clementson

2 1

Re: [slime-devel] Automatic "semantic" indentation
by Julian Stecklina 05 Apr '04

05 Apr '04

Luke Gorrie <luke(a)bluetail.com> writes: > Ahoy, > > I added a hopefully nice feature: Lisp teaches Emacs how to indent > macros. I found a way to do it without hacking Emacs's indentation > code at all. > >>From now on when you have a macro like: > > (defmacro foo (a b &body body) > ...) > > anywhere in your Lisp image, Emacs should know to indent calls as: > > (foo ((x y)) > (bar) > (baz) > (beer)) > > i.e. it groks &body. This is a very nice thing to have! Regards, -- Julian Stecklina Signed and encrypted mail welcome. Key-Server: pgp.mit.edu Key-ID: 0xD65B2AB5 FA38 DCD3 00EC 97B8 6DD8 D7CC 35D8 8D0E D65B 2AB5 Any sufficiently complicated C or Fortran program contains an ad hoc informally-specified bug-ridden slow implementation of half of Common Lisp. - Greenspun's Tenth Rule of Programming

1 0

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 05 Apr '04

05 Apr '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.320 slime/ChangeLog:1.321 --- slime/ChangeLog:1.320 Thu Apr 1 17:21:23 2004 +++ slime/ChangeLog Mon Apr 5 02:27:48 2004 @@ -1,7 +1,44 @@ -2004-04-02 Lawrence Mitchell <wence(a)gmx.li> +2004-04-05 Lawrence Mitchell <wence(a)gmx.li> - * slime.el (slime-autodoc): Escape %-signs in arglist before - passing to format. + * swank.lisp (*swank-pprint-circle*, *swank-pprint-escape*, + *swank-pprint-level*, *swank,pprint-length*): Fix typo in + docstring. + + * slime.el (slime-arglist): Don't `message' arglist directly, in + case it contains %-signs. + (slime-repl-output-face): Fix quoting. + (slime-symbol-at-point): Call `slime-symbol-name-at-point', + rather than ourselves. + (slime-check-protocol-version): Docstring fix. + +2004-04-05 Luke Gorrie <luke(a)bluetail.com> + + * doc/slime.texi (Semantic indentation): Documented new + automatically-learn-how-to-indent-macros feature. + Added auto version control header in subtitle. + + * slime.el (slime-close-parens-at-point): New command bound to + C-a C-a. Inserts close-parenthesis characters at point until the + top-level form becomes well formed. Could perhaps be made fancier. + (slime-update-indentation): New command to update indentation + information (`common-lisp-indent-function' properties) based on + macro information extracted from Lisp. This happens + automatically, the command is just to force a full rescan. + + * swank.lisp (connection): Added slots to track indentation caching. + (*connections*): List of all open connections. + (default-connection): Function to get a "default" + connection. This is intended to support globally using the + debugger hook outside the context of a SLIME request, which is + broken at present. + (with-connection): Don't setup a restart: that must be done + separately. + (sync-state-to-emacs): Call `update-connection-indentation'. + (update-connection-indentation): Automatically discover how to + indent macros and tell Emacs. + + * swank-backend.lisp (arglist): Specify that strings returned + from ARGLIST should be READable. 2004-04-02 Helmut Eller <e9626484(a)stud3.tuwien.ac.at>

1 0

[slime-devel] Bad interaction between slime and repl variables in SBCL
by Lawrence Mitchell 05 Apr '04

05 Apr '04

Using the latest version of slime (2004-04-02), and SBCL 0.8.9.6, I get problems with repl variables. To whit: ; SLIME 2004-04-02 CL-USER> (find-package "CL") #<PACKAGE "COMMON-LISP"> CL-USER> * T CL-USER> + (LOAD "/opt/cvs-trees/slime/swank-loader") CL-USER> I'm not sure where to start looking for problems here, since adding FORMAT statements in LISTENER-EVAL seems to set *, **, and their brethren correctly. Evaluating the same forms in a command-line SBCL give's me the expected results: > (find-package "CL") #<PACKAGE "COMMON-LISP"> > * #<PACKAGE "COMMON-LISP"> > + * > -- Lawrence Mitchell <wence(a)gmx.li>

2 2

[slime-devel] A few minor fixes.
by Lawrence Mitchell 05 Apr '04

05 Apr '04

Here are a few fixes which are sitting around in my local slime files. There's also a fix for the ChangeLog entry regarding slime-autodoc-mode, which Helmut fixed in a cleaner way than me, the entry now refers to the changed function. Index: ChangeLog =================================================================== RCS file: /project/slime/cvsroot/slime/ChangeLog,v retrieving revision 1.320 diff -u -r1.320 ChangeLog --- ChangeLog 1 Apr 2004 22:21:23 -0000 1.320 +++ ChangeLog 2 Apr 2004 21:48:36 -0000 @@ -1,7 +1,15 @@ 2004-04-02 Lawrence Mitchell <wence(a)gmx.li> - * slime.el (slime-autodoc): Escape %-signs in arglist before - passing to format. + * swank.lisp (*swank-pprint-circle*, *swank-pprint-escape*, + *swank-pprint-level*, *swank,pprint-length*): Fix typo in + docstring. + + * slime.el (slime-arglist): Don't `message' arglist directly, in + case it contains %-signs. + (slime-repl-output-face): Fix quoting. + (slime-symbol-at-point): Call `slime-symbol-name-at-point', + rather than ourselves. + (slime-check-protocol-version): Docstring fix. 2004-04-02 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.250 diff -u -r1.250 slime.el --- slime.el 1 Apr 2004 22:20:14 -0000 1.250 +++ slime.el 2 Apr 2004 21:40:13 -0000 @@ -180,7 +180,7 @@ :group 'slime) (defface slime-repl-output-face - `((t (:inherit font-lock-string-face))) + '((t (:inherit font-lock-string-face))) "Face for Lisp output in the SLIME REPL." :group 'slime) @@ -818,7 +818,7 @@ (defun slime-symbol-at-point () "Return the symbol at point, otherwise nil." - (let ((name (slime-symbol-at-point))) + (let ((name (slime-symbol-name-at-point))) (and name (intern name)))) (defun slime-sexp-at-point () @@ -1125,7 +1125,7 @@ (setq slime-changelog-date (slime-changelog-date)) (defun slime-check-protocol-version (lisp-version) - "Signal an error LISP-VERSION equal to `slime-changelog-date'" + "Signal an error unless LISP-VERSION is equal to `slime-changelog-date'." (unless (and lisp-version (equal lisp-version slime-changelog-date)) (slime-disconnect) (let ((message (format "Protocol mismatch: Lisp: %s ELisp: %s" Index: swank.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank.lisp,v retrieving revision 1.156 diff -u -r1.156 swank.lisp --- swank.lisp 30 Mar 2004 23:08:31 -0000 1.156 +++ swank.lisp 2 Apr 2004 21:42:37 -0000 @@ -1086,16 +1086,16 @@ (prin1-to-string (eval form)))))) (defvar *swank-pprint-circle* *print-circle* - "*PRINT-CIRCLE* is bound to this volue when pretty printing slime output.") + "*PRINT-CIRCLE* is bound to this value when pretty printing slime output.") (defvar *swank-pprint-escape* *print-escape* - "*PRINT-ESCAPE* is bound to this volue when pretty printing slime output.") + "*PRINT-ESCAPE* is bound to this value when pretty printing slime output.") (defvar *swank-pprint-level* *print-level* - "*PRINT-LEVEL* is bound to this volue when pretty printing slime output.") + "*PRINT-LEVEL* is bound to this value when pretty printing slime output.") (defvar *swank-pprint-length* *print-length* - "*PRINT-LENGTH* is bound to this volue when pretty printing slime output.") + "*PRINT-LENGTH* is bound to this value when pretty printing slime output.") (defun swank-pprint (list) "Bind some printer variables and pretty print each object in LIST." -- Lawrence Mitchell <wence(a)gmx.li>

2 1

[slime-devel] Automatic "semantic" indentation
by Luke Gorrie 05 Apr '04

05 Apr '04

Ahoy, I added a hopefully nice feature: Lisp teaches Emacs how to indent macros. I found a way to do it without hacking Emacs's indentation code at all. >From now on when you have a macro like: (defmacro foo (a b &body body) ...) anywhere in your Lisp image, Emacs should know to indent calls as: (foo ((x y)) (bar) (baz) (beer)) i.e. it groks &body. Details and caveats are in the manual (and I put fresh versions in my ~/misc/lisp web directory). Feedback / tweak-suggestions appreciated. -Luke

1 0

[slime-devel] [PATCH]: put working directory in header-line
by Sean O'Rourke 05 Apr '04

05 Apr '04

This minor patch compresses the existing header-line information in the REPL, and adds the current working directory (*default-pathname-defaults*). I think it's a nice bit of information to have there when you're dealing with files from multiple directories. Do other people find it useful? /s

1 0

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 02 Apr '04

02 Apr '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.318 slime/ChangeLog:1.320 --- slime/ChangeLog:1.318 Wed Mar 31 17:46:55 2004 +++ slime/ChangeLog Thu Apr 1 17:21:23 2004 @@ -1,13 +1,33 @@ +2004-04-02 Lawrence Mitchell <wence(a)gmx.li> + + * slime.el (slime-autodoc): Escape %-signs in arglist before + passing to format. + +2004-04-02 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * slime.el (slime-maybe-list-compiler-notes): Display the notes + for C-c C-c, when there are notes without a good source-location. + +2004-04-01 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank-sbcl.lisp: Remove the non-working workarounds for + non-existent fcntl. Reported by Brian Mastenbrook. + (preferred-communication-style): Use multithreading if futexes are + available, sigio if fcntl is present, and fd-handlers otherwise. + (resolve-note-location): Don't try to construct a source-location + if there's no context. Notes without location will be displayed + in the note-listing buffer. + +2004-04-01 Bill Clementson <Bill_Clementson(a)peoplesoft.com> + + * swank-allegro.lisp (send): Fix misplaced parens. + 2004-03-31 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank-cmucl.lisp (debug-function-arglist): Return symbols if possible. (class-location): Support for experimental source-location recording. - -2004-04-01 Bill Clementson <Bill_Clementson(a)peoplesoft.com> - - * swank-allegro.lisp (send): Fix misplaced parens. 2004-03-30 Helmut Eller <e9626484(a)stud3.tuwien.ac.at>

1 0

[slime-devel] fix autodoc for functions with %-signs in their arglists
by Lawrence Mitchell 01 Apr '04

01 Apr '04

Occasionally, I get errors when using slime-autodoc-mode, due to arglists having %-signs in them. This patch should fix the problem. Index: ChangeLog =================================================================== RCS file: /project/slime/cvsroot/slime/ChangeLog,v retrieving revision 1.319 diff -u -r1.319 ChangeLog --- ChangeLog 1 Apr 2004 21:35:10 -0000 1.319 +++ ChangeLog 1 Apr 2004 22:07:38 -0000 @@ -1,3 +1,8 @@ +2004-04-01 Lawrence Mitchell <wence(a)gmx.li> + + * slime.el (slime-autodoc): Escape %-signs in arglist before + passing to format. + 2004-04-01 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank-sbcl.lisp: Remove the non-working workarounds for Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.249 diff -u -r1.249 slime.el --- slime.el 30 Mar 2004 23:10:46 -0000 1.249 +++ slime.el 1 Apr 2004 22:05:36 -0000 @@ -3223,6 +3223,9 @@ (slime-buffer-package) (with-lexical-bindings (cache-key name) (lambda (arglist) + (let ((start 0)) + (while (setq start (1+ (string-match "%" arglist start))) + (setq arglist (replace-match "%%" nil t arglist)))) ;; FIXME: better detection of "no documentation available" (if (string-match "<not available>" arglist) (setq arglist "")) -- Lawrence Mitchell <wence(a)gmx.li>

2 1

[slime-devel] SBCL swank defconstants
by Brian Mastenbrook 01 Apr '04

01 Apr '04

Howdy, The other day I attempted to use CVS SLIME to speak to a SBCL running on Solaris via fd-handlers, and found that there are some defconstant forms which are fouling things up anyway: (defconstant +o_async+ #+linux 8192 #+bsd #x40) (defconstant +f_setown+ #+linux 8 #+bsd 6) (defconstant +f_setfl+ #+(or linux bsd) 4) Each of these should probably have a #-(linux bsd) -1 form or something like that. SIGIO seems to disagree with SBCL on OS X as well, so I'm back to using fd-handlers there too. Hope that option doesn't go away! Brian -- Brian Mastenbrook bmastenb(a)cs.indiana.edu http://cs.indiana.edu/~bmastenb/

2 1