March 2006 - slime-devel - mailman3.common-lisp.net

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 26 Mar '06

26 Mar '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.870 slime/ChangeLog:1.871 --- slime/ChangeLog:1.870 Fri Mar 24 02:30:34 2006 +++ slime/ChangeLog Sat Mar 25 22:58:16 2006 @@ -1,3 +1,26 @@ +2006-03-25 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + * swank.lisp (arglist-for-echo-area): New keyword arg, + print-lines. + (decoded-arglist-to-string): New function, implement argument + highlighting also for &optional and &rest/&body arguments. + (arglist-to-string): Use decoded-arglist-to-string. + (arglist): New slots aux-args, known-junk, unknown-junk. + (nreversef): New macro. + (decode-arglist, encode-arglist): Refine to handle more structure + in argument lists, including implementation-defined stuff like + &parse-body. + (format-arglist-for-echo-area): New keyword arg, print-lines. + Simplify the code as there is no need to fall back to the unparsed + arglist any more. + + * slime.el (slime-fontify-string): Fix for arguments spanning + multiple lines. + (slime-autodoc-message-dimensions): New. + (slime-autodoc-thing-at-point): Use it here to either ask for a + one-line or a nicely formatted multi-line arglist. + (slime-enclosing-operator-names): Handle linebreaks. + 2006-03-24 Mikel Bancroft <mutandiz(a)yahoo.com> * swank-allegro.lisp (set-default-directory): Fix for pathnames

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 25 Mar '06

25 Mar '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.869 slime/ChangeLog:1.870 --- slime/ChangeLog:1.869 Thu Mar 23 18:28:07 2006 +++ slime/ChangeLog Fri Mar 24 02:30:34 2006 @@ -1,3 +1,8 @@ +2006-03-24 Mikel Bancroft <mutandiz(a)yahoo.com> + + * swank-allegro.lisp (set-default-directory): Fix for pathnames + without a trailing slash. + 2006-03-24 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> * slime.el (slime-background-activities-enabled-p): Allow

1 0

[slime-devel] AllegroCL and ,cd without a trailing slash
by Mikel Bancroft 24 Mar '06

24 Mar '06

I've run into a problem in some emacs versions where changing directories via ,cd without a trailing slash will cause subsequent cd commands (and possibly other pathname related lisp routines) to fail. The problem is that read-directory-name may or may not return a path string with a trailing slash. So a directory "/users/mikel/foo" that is passed to set-default-directory in the allegro backend and is parsed into a pathname with 'foo' in the :name slot rather than as part of the :directory. This is then stored into *default-pathname-defaults*. The following patch avoids this since the path returned by excl:chdir returns a path string with a trailing slash regardless of whether or not its argument does and thus will be parsed correctly when converted into a pathname object. Index: swank-allegro.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank-allegro.lisp,v retrieving revision 1.85 diff -c -r1.85 swank-allegro.lisp *** swank-allegro.lisp 22 Mar 2006 16:40:01 -0000 1.85 --- swank-allegro.lisp 24 Mar 2006 06:45:05 -0000 *************** *** 86,94 **** "allegro") (defimplementation set-default-directory (directory) ! (let ((dir (namestring (setf *default-pathname-defaults* ! (truename (merge-pathnames directory)))))) ! (excl:chdir dir) dir)) (defimplementation default-directory () --- 86,93 ---- "allegro") (defimplementation set-default-directory (directory) ! (let* ((dir (namestring (truename (merge-pathnames directory))))) ! (setf *default-pathname-defaults* (pathname (excl:chdir dir))) dir)) (defimplementation default-directory () -Mikel Bancroft

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 24 Mar '06

24 Mar '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.867 slime/ChangeLog:1.869 --- slime/ChangeLog:1.867 Thu Mar 23 00:02:54 2006 +++ slime/ChangeLog Thu Mar 23 18:28:07 2006 @@ -1,3 +1,29 @@ +2006-03-24 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + * slime.el (slime-background-activities-enabled-p): Allow + "background activities" in sldb-mode. + (slime-autodoc-message-ok-p): Allow autodoc in sldb-mode. + (sldb-mode-syntax-table): New variable. + (sldb-mode): Enable autodoc-mode when slime-use-autodoc-mode is + true. Use sldb-mode-syntax-table to make #<...> balance like + parentheses. This enables autodoc-mode to match #<unreadable> + actual arguments in the backtraces with formal arguments of the + function. + (slime-beginning-of-symbol, slime-end-of-symbol): Handle + es::|caped| symbols. + (slime-enclosing-operator-names): Use syntax table to check + whether we are at the beginning of a balanced expression. + +2006-03-23 Christophe Rhodes <csr21(a)cam.ac.uk> + + * swank.lisp (ed-in-emacs): Allow conses as function names. + Ensure that there is a connection to emacs before sending the + :ed message. + + * slime.el (slime-edit-definition): read names, not symbols. + (slime-ed): handle conses whose car is not a string as function + names. + 2006-03-23 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> * slime.el (slime-qualify-cl-symbol-name): Strip leading colon

1 0

[slime-devel] repl, stdio flush mode, sbcl started from a core file
by Attila Lendvai 23 Mar '06

23 Mar '06

hi! did anybody notice that when using slime with sbcl starting from a custom core file, then the communication channel between the repl buffer and sbcl seems to be buffered? (sometimes i need to press an enter for the ucw log output or compile output to arrive in emacs) i don't seem to experience this if i start a vanilla sbcl and (require ...) the fasl's into it, but it takes ages compared to starting from a core file. i create my sbcl core file from xemacs, so swank is already loaded when the core file is saved, and my guess is that when the communication channel is re-estabilished between slime and swank, then the flush mode on the unix file handle is not restored properly. (i have no experience with how sbcl handles external resources) unfortunately i can't test right now what happens when not loading swank into the core file, but i'll try it tonight. - attila (alias 101 on irc &no 'its not lisp code :)

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 23 Mar '06

23 Mar '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.863 slime/ChangeLog:1.867 --- slime/ChangeLog:1.863 Tue Mar 21 21:48:38 2006 +++ slime/ChangeLog Thu Mar 23 00:02:54 2006 @@ -1,5 +1,56 @@ +2006-03-23 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + * slime.el (slime-qualify-cl-symbol-name): Strip leading colon + from package names for qualifying symbols. + (slime-call-defun): New command. + (slime-keys): Bind it to C-c C-y. + (slime-easy-menu): Show it in the menu. + + * slime.el (slime-autodoc-use-multiline-p): New defcustom. + (slime-autodoc-message): Use it here. Fix bug that autodoc + messages exceeding one line could not be overwritten by later + autodoc messages. + (slime-autodoc-pre-command-refresh-echo-area): Use message + rather than slime-background-message. + + * swank.lisp (casify): Removed. + (casify-char, tokenize-symbol-thoroughly): New functions. + (parse-symbol): Use tokenize-symbol-thoroughly, so as to handle + |escaped symbols|. This fixes arglist display for operators with + strange symbol names. + +2006-03-23 Douglas Crosher <dcrosher(a)common-lisp.net> + + * swank-backend (accept-connection): add a 'timeout argument to + this function. + + * swank-backend (set-stream-timeout): new implementation specific + function. Used to set the timeout for stream operations, which + can help make the network connection establishment more robust. + + * swank (setup-server): ignore errors from the function 'serve to + allow another connection to be made. + + * swank (serve-connection): ensure the listener socket is closed + when 'dont-close is false, even if the connection attempt fails. + + * swank (accept-authenticated-connection): ensure the new + connection is closed if the connection establishment fails. Set a + short stream timeout to prevent denial of survice. + + * swank (open-dedicated-output-stream): ensure the listener socket + is closed, even if unable to open the dedicated stream. Implement + a timeout while waiting for a connection for the dedicate stream + to prevent denial of service. + + * swank (create-connection): ensure the new connection is closed + if not successful. + 2006-03-22 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + * swank.lisp (arglist-for-echo-area): Fix when arg-indices are + not given. + * slime.el (slime-ed): Handle (FILENAME :charpos CHARPOS). * swank.lisp (inspect-for-emacs): Specialize on FILE-STREAM and

1 0

[slime-devel] Propose a patch to fix load error on Win32
by Pisin Bootvong 22 Mar '06

22 Mar '06

Hi, This patch should fix source file not found error when: - Run on Win32 (CLISP) - (user-homedir-pathname) is on different drive than *default-pathname-defaults* - Run slime a second time (so slime has to compare fasl file date with source code date). The problem is that PATHNAME-DIRECTORY does not include the drive letter. Please review and apply if it is ok. --- swank-loader-cvs-20060311.lisp 2006-02-25 21:57:21.000000000 +0700 +++ swank-loader.lisp 2006-03-11 16:27:18.625000000 +0700 @@ -179,8 +179,9 @@ (defun swank-source-files (source-directory) (mapcar (lambda (name) - (make-pathname :name name :type "lisp" - :directory (pathname-directory source-directory))) + (merge-pathnames + (make-pathname :name name :type "lisp") + source-directory)) `("swank-backend" ,@*sysdep-files* "swank"))) (defvar *fasl-directory* (default-fasl-directory)

1 0

[slime-devel] tramp-make-tramp-file-name
by Marco Baringer 21 Mar '06

21 Mar '06

my version of tramp-make-tramp-file-name (emacs 22.0.50.1) takes 5 arguments, not 4 as in slime-create-filename-translator. is this a bug or does this function have a differenet api on other emacs? -- -Marco Ring the bells that still can ring. Forget the perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen

4 3

[slime-devel] slime in its own frame?
by Rick Taube 21 Mar '06

21 Mar '06

how can i reliably start the slime repl running in its own dedicated frame? this little code snippet does open a new frame and start slime running inside it with the new slime frame the active frame. however, if i launch the new frame and then RESELECT my original frame with the mouse while "*inferior-lisp*" is booting but before the "*slime repl*" actually starts running, then the "*slime repl*" actually gets opened back in the original frame. that is, i want the "*slime repl*" to start in whatever frame the *inferior-lisp* buffer was started in, not whatever the currently active frame happens to be. (defun slime-own-frame () (let* ((inf "*inferior lisp*")) (switch-to-buffer-other-frame inf) (with-current-buffer inf (slime)) (bury-buffer inf))) --rick

1 0

Re: [slime-devel] Re: Problems getting {}[] recognized as parenthesis
by jjkola＠email.com 19 Mar '06

19 Mar '06

"GP lisper" <spambait(a)CloudDancer.com> wrote: > One night on IRC, antifuchs worked out this for me: > (add to your emacs defaults) > > ; fix [] in slime > (add-hook 'lisp-mode-hook > (lambda () > (slime-mode t) > (modify-syntax-entry ?\[ "(] " lisp-mode-syntax-table) > (modify-syntax-entry ?\] ")[ " lisp-mode-syntax-table))) It worked like a charm. Thanks a lot. Jyrki Jaakkola

1 0