June 2005 - slime-devel - mailman3.common-lisp.net

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 28 Jun '05

28 Jun '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.716 slime/ChangeLog:1.717 --- slime/ChangeLog:1.716 Tue Jun 21 20:28:58 2005 +++ slime/ChangeLog Tue Jun 28 10:40:07 2005 @@ -1,3 +1,8 @@ +2005-06-28 Gabor Melis <mega(a)hotpop.com> + + * swank-sbcl.lisp (threaded stuff): horrible hack to make threaded + SBCL 0.9.2 work. (also, Happy Birthday Christophe!) + 2005-06-21 Edi Weitz <edi(a)agharta.de> * swank.lisp (find-matching-packages): Also use nicknames.

1 0

Re: [slime-devel] Re: [Openmcl-devel] C-c C-c of a function refer encing a large constant takes a long time to compile
by Hoehle, Joerg-Cyril 27 Jun '05

27 Jun '05

Hi, David Steuber wrote: >SLIME or Swank (perhaps the swank-openmcl backend) are definitely doing >something to make this take a long time. I just have no idea how to >figure out what. I suspect some interaction between EVAL-WHEN and top-level vs. non top-level. I've also had a couple of situations (in CLISP, cmucl or sbcl, but I forgot which package and when) where e.g. C-c C-c would not work and even spit out incomprehensible error messages, while e.g. C-x C-e followed by (compile 'foo) would work fine. My belief is the difference that evaluating (funcall (compile nil '(lambda() (defun foo ...)))) -- which is what I believe SLIME does (cf. swank-compile-string) -- may have different results from evaluating 1. (defun foo) -- causes it to be defined at top-level 2. (compile 'foo) Look at the CLHS apropos EVAL-WHEN and you'll see the door open to a lot of hairy behaviour. People mostly using COMPILE-FILE on lots of top-level definitions will never encounter that, because it's the typical, well-tested use-case. The above is not. You'll be hit by these differences when you'll see non-trivial EVAL-WHEN ... in the macroexpansions of the forms you evaluate or compile individually in SLIME. Sigh, I wish I had kept notes about those non-top-level problems, to beter talk about these. BTW, IIRC, there's a few situations where the old lisp-mode (inf-lisp.el) is less likely to run into trouble than SLIME, precisely because it just uses a simpler way to compilation and evaluation. Again, I forgot the details. Also, I've wondered whether platform-specific solutions may help. E.g. in CLISP, (locally #+clisp (declare (compile)) ...whatever...) might be a better approach to compilation, devoid of top-level vs. non top-level EVAL-WHEN expansion issues. CLHS specifies that LOCALLY maintains top-levelness, but (declare compile) is specific to CLISP. Actually, I think this could be an excellent idea for swank-clisp:defimplementation swank-compile-string. Or just use (load (make-string-input-stream "(print 123)"):compiling t :verbose nil) [looking closer at swank-*.lisp] Well forget about my talking, it appears only clisp uses (funcall(compile(lambda #))). Allegro, Lispworks, sbcl, openmcl go through a temp file, and cmucl uses ext:compile-from-stream on a string-input-stream, at least in slime-1.2.1. Regards, Jorg Hohle.

1 0

Re: [slime-devel] feture suggjestion
by Knut Olav Bøhmer 27 Jun '05

27 Jun '05

OK, then I'll try to implement it, when I get time. (which will not be before the end of July, if not then, then next year. I hope someone has more time for this than I :) killing threads in a infinite loop is like killing any other thread. slime-list-threads (C-c C-x t) then arrow up and down to your (un)favorite thread of choice, then press "k". On 6/23/05, Jeffrey Cunningham <jeffrey(a)cunningham.net> wrote: > I like it. Incidentally, how do you kill any thread caught in an > infinite loop? I've been trying to figure this out for sometime. > > --Jeff Cunningham > > > > On Thu Jun 23, 2005 at 12:02:06PM +0200, Knut Olav B?hmer wrote: > > When I'm in a buffer editing my lisp-code, and decide to test the > > code, by pressing C-x C-e, slime forks of a new thread with the name > > "worker". The problem starts when I have several threads running from > > different buffers with the name "worker", and one of the threads ends > > in an infinite loop or deadlock, and I want to kill one of the > > threads. Then I don't know which thread is doing what. > > > > So If the thread got the name "worker + buffer name" then I would have > > slightly more control. > > > > Good Idea? > > > > > > -- > > Knut Olav Bøhmer > > tlf: +47 47400777 > > _______________________________________________ > > slime-devel site list > > slime-devel(a)common-lisp.net > > http://common-lisp.net/mailman/listinfo/slime-devel > > > --Jeff > > -- Knut Olav Bøhmer tlf: +47 47400777

3 5

[slime-devel] fuzzy completion code
by PERROT Stephane 24 Jun '05

24 Jun '05

Hello all slimers, I would like to copy out swank.lisp fuzzy-completion code in a distinct file, so as to be able do distribute it independantly of SLIME, for use in LispWorks Editor (as people using only LW Editor don't necessarily have SLIME installed) Has anybody an objection to me doing that? Also, I would like to know who the author(s) of this part of swank.lisp was (were), in order to give credit where credit is due... Thanks in advance for your answer, Stéphane Perrot P.S. : In my request you can see the result of Luke's brilliant Amsterdam SLIME demo, where all users of other Lisp environnments instantly became envious of that marvelous, unheard-of, feature (fuzzy completion) :-)

2 1

[slime-devel] Patch to improve presentations
by Matthias Koeppe 23 Jun '05

23 Jun '05

Hi, thanks to all who have worked on the new presentations feature of SLIME. This is very nice work! I am sending a patch that make the presentations feature a bit more robust and intuitive (IMHO). With the patch, parts of presentations can be copied reliably using all available Emacs facilities (not just kill-ring-save), and they are no longer "semi-readonly" (in the sense that keypresses are silently ignored). Whenever a user attempts to edit a presentation, it now simply turns into plain text (which is indicated by changing the face); this can be undone. The patch removes the pre-command and post-command hooks and the classification of some modification commands into "action-type"s. In an after-change-function, I check whether only a part of a presentation has been pasted or whether a presentation has been edited. Cheers, Matthias Index: slime.el =================================================================== RCS file: /project/slime/cvsroot/slime/slime.el,v retrieving revision 1.509 diff -u -p -u -r1.509 slime.el --- slime.el 12 Jun 2005 21:05:29 -0000 1.509 +++ slime.el 23 Jun 2005 14:57:12 -0000 @@ -860,15 +860,13 @@ This list of flushed between commands.") "Execute all functions in `slime-pre-command-actions', then NIL it." (dolist (undo-fn slime-pre-command-actions) (ignore-errors (funcall undo-fn))) - (setq slime-pre-command-actions nil) - (slime-presentation-command-hook)) + (setq slime-pre-command-actions nil)) (defun slime-post-command-hook () (when (and slime-mode (slime-connected-p)) (slime-process-available-input)) (when (null pre-command-hook) ; sometimes this is lost - (add-hook 'pre-command-hook 'slime-pre-command-hook)) - (slime-presentation-post-command-hook) ) + (add-hook 'pre-command-hook 'slime-pre-command-hook))) (defun slime-setup-command-hooks () "Setup a buffer-local `pre-command-h'ook' to call `slime-pre-command-hook'." @@ -876,7 +874,8 @@ This list of flushed between commands.") (make-local-hook 'post-command-hook) ;; alanr: need local t (add-hook 'pre-command-hook 'slime-pre-command-hook nil t) - (add-hook 'post-command-hook 'slime-post-command-hook nil t)) + (add-hook 'post-command-hook 'slime-post-command-hook nil t) + (add-hook 'after-change-functions 'slime-after-change-function nil t)) ;(add-hook 'slime-mode-hook 'slime-setup-command-hooks) ;(setq post-command-hook nil) @@ -2560,6 +2559,8 @@ update window-point afterwards. If poin (when (boundp 'text-property-default-nonsticky) (pushnew '(slime-repl-old-output . t) text-property-default-nonsticky :test 'equal) + (pushnew '(slime-repl-presentation . t) text-property-default-nonsticky + :test 'equal) (pushnew '(slime-repl-result-face . t) text-property-default-nonsticky :test 'equal))) @@ -2584,16 +2585,56 @@ update window-point afterwards. If poin (setf (gethash id slime-presentation-start-to-point) nil) (when start (with-current-buffer (slime-output-buffer) - (add-text-properties - start (symbol-value 'slime-output-end) - `(face slime-repl-result-face - slime-repl-old-output ,id - mouse-face slime-repl-output-mouseover-face - keymap ,slime-presentation-map - rear-nonsticky (slime-repl-old-output - slime-repl-result-face - slime-repl-output-mouseover-face)))))))))) + (slime-add-presentation-properties start (symbol-value 'slime-output-end) + id nil)))))))) + +(defstruct (slime-presentation) + (text) + (id) + (start-p) + (stop-p)) + +(defun slime-add-presentation-properties (start end id result-p) + "Make the text between START and END a presentation with ID. +RESULT-P decides whether a face for a return value or output text is used." + (add-text-properties start end + `(face slime-repl-inputed-output-face + slime-repl-old-output ,id + mouse-face slime-repl-output-mouseover-face + keymap ,slime-presentation-map + rear-nonsticky (slime-repl-old-output + slime-repl-presentation + slime-repl-result-face + slime-repl-output-mouseover-face))) + (let ((text (buffer-substring-no-properties start end))) + (case (- end start) + (0) + (1 + (add-text-properties start end + `(slime-repl-presentation + ,(make-slime-presentation :text text :id id + :start-p t :stop-p t)))) + (t + (let ((inhibit-modification-hooks t)) + (add-text-properties start (1+ start) + `(slime-repl-presentation + ,(make-slime-presentation :text text :id id + :start-p t :stop-p nil))) + (when (> (- end start) 2) + (add-text-properties (1+ start) (1- end) + `(slime-repl-presentation + ,(make-slime-presentation :text text :id id + :start-p nil :stop-p nil)))) + (add-text-properties (1- end) end + `(slime-repl-presentation + ,(make-slime-presentation :text text :id id + :start-p nil :stop-p t)))))))) +(defun slime-insert-presentation (result output-id) + (let ((start (point))) + (insert result) + (slime-add-presentation-properties start (point) (- output-id) t))) + (defun slime-open-stream-to-lisp (port) (let ((stream (open-network-stream "*lisp-output-stream*" (slime-with-connection-buffer () @@ -2746,61 +2787,105 @@ joined together.")) (slime-setup-command-hooks) (run-hooks 'slime-repl-mode-hook)) -(defvar slime-not-copying-whole-presentation nil) - -;; alanr -(defun slime-presentation-command-hook () - (let* ((props-here (text-properties-at (point))) - (props-before (and (not (= (point) (point-min))) - (text-properties-at (1- (point))))) - (inside (and (getf props-here 'slime-repl-old-output))) - (at-beginning (and inside - (not (getf props-before 'slime-repl-old-output)))) - (at-end (and (or (= (point) (point-max)) - (not (getf props-here 'slime-repl-old-output))) - (getf props-before 'slime-repl-old-output))) - (start (cond (at-beginning (point)) - (inside (previous-single-property-change - (point) 'slime-repl-old-output)) - (at-end (previous-single-property-change - (1- (point)) 'slime-repl-old-output)))) - (end (cond (at-beginning (or (next-single-property-change - (point) 'slime-repl-old-output) - (point-max))) - (inside (or (next-single-property-change (point) 'slime-repl-old-output) - (point-max))) - (at-end (point))))) - ; (setq message (format "%s %s %s %s %s" at-beginning inside at-end start end)) - (when (and (or inside at-end) start end (> end start)) - (let ((kind (get this-command 'action-type))) - ; (message (format "%s %s %s %s" at-beginning inside at-end kind)) - (cond ((and (eq kind 'inserts) inside (not at-beginning)) - (setq this-command 'ignore)) - ((and (eq kind 'deletes-forward) inside (not at-end)) - (kill-region start end) - (setq this-command 'ignore)) - ((and (eq kind 'deletes-backward) (or inside at-end) (not at-beginning)) - (kill-region start end) - (setq this-command 'ignore)) - ((eq kind 'copies) - (multiple-value-bind (start end) (slime-property-bounds 'slime-repl-old-input) - (setq slime-not-copying-whole-presentation - (not (or (and at-beginning (>= (mark) end)) - (and at-end (<= (mark) start))))))) - ;(message (format "%s %s" length (abs (- (point) (mark)))))))) - ))))) - -;; if we did not copy the whole presentation, then remove the text properties from the -;; top of the kill ring - -(defun slime-presentation-post-command-hook () - (when (eq (get this-command 'action-type) 'copies) - (when slime-not-copying-whole-presentation - (remove-text-properties 0 (length (car kill-ring)) - '(slime-repl-old-output t mouse-face t rear-nonsticky t) - (car kill-ring)))) - (setq slime-not-copying-whole-presentation nil) - ) +(defun slime-presentation-whole-p (start end) + (let ((presentation (get-text-property start 'slime-repl-presentation))) + (and presentation + (string= (buffer-substring-no-properties start end) + (slime-presentation-text presentation))))) + +(defun slime-same-presentation-p (a b) + (and (string= (slime-presentation-text a) (slime-presentation-text b)) + (= (slime-presentation-id a) (slime-presentation-id b)))) + +(defun* slime-presentation-start () + "Find start of presentation at point. Return buffer index and + whether a start-tag was found. When there is no presentation at + point, return nil and nil." + (let* ((presentation (get-text-property (point) 'slime-repl-presentation)) + (this-presentation presentation)) + (unless presentation + (return-from slime-presentation-start + (values nil nil))) + (save-excursion + (while (not (slime-presentation-start-p this-presentation)) + (let ((change-point (previous-single-property-change (point) 'slime-repl-presentation))) + (unless change-point + (return-from slime-presentation-start + (values (point-min) nil))) + (setq this-presentation (get-text-property change-point 'slime-repl-presentation)) + (unless (and this-presentation + (slime-same-presentation-p presentation this-presentation)) + (return-from slime-presentation-start + (values (point) nil))) + (goto-char change-point))) + (values (point) t)))) + +(defun* slime-presentation-end () + "Find end of presentation at point. Return buffer index (after last + character of the presentation) and whether an end-tag was found." + (let* ((presentation (get-text-property (point) 'slime-repl-presentation)) + (this-presentation presentation)) + (unless presentation + (return-from slime-presentation-end + (values nil nil))) + (save-excursion + (while (and this-presentation + (slime-same-presentation-p presentation this-presentation) + (not (slime-presentation-stop-p this-presentation))) + (let ((change-point (next-single-property-change (point) 'slime-repl-presentation))) + (unless change-point + (return-from slime-presentation-end + (values (point-max) nil))) + (goto-char change-point) + (setq this-presentation (get-text-property (point) 'slime-repl-presentation)))) + (if (and this-presentation + (slime-same-presentation-p presentation this-presentation)) + (let ((after-end (next-single-property-change (point) 'slime-repl-presentation))) + (if (not after-end) + (values (point-max) t) + (values after-end t))) + (values (point) nil))))) + +(defun slime-presentation-around-point () + "Return start index, end index, and whether the presentation is complete." + (multiple-value-bind (start good-start) + (slime-presentation-start) + (multiple-value-bind (end good-end) + (slime-presentation-end) + (values start end + (and good-start good-end + (slime-presentation-whole-p start end)))))) + +(defun slime-after-change-function (start end old-len) + "Check all presentations within and adjacent to the change. When a + presentation has been altered, change it to plain text." + (unless undo-in-progress + (let ((real-start (max (point-min) (1- start))) + (real-end (min (point-max) (1+ end))) + (any-change nil)) + ;; positions around the change + (save-excursion + (goto-char real-start) + (while (< (point) real-end) + (let ((presentation (get-text-property (point) 'slime-repl-presentation))) + (when presentation + (multiple-value-bind (from to whole) + (slime-presentation-around-point) + ;;(message "presentation %s whole-p %s" (buffer-substring from to) whole) + (unless whole + (setq any-change t) + (remove-text-properties from to + '(slime-repl-old-output t + slime-repl-inputed-output-face t + face t mouse-face t rear-nonsticky t + slime-repl-presentation t)))))) + (let ((next-change + (next-single-property-change (point) 'slime-repl-presentation nil + real-end))) + (if next-change + (goto-char next-change) + (undo-boundary) + (return)))))))) (defun slime-copy-presentation-at-point (event) (interactive "e") @@ -2824,20 +2909,6 @@ joined together.")) (goto-char (point-max)) (do-insertion))))))) -(put 'self-insert-command 'action-type 'inserts) -(put 'self-insert-command-1 'action-type 'inserts) -(put 'yank 'action-type 'inserts) -(put 'kill-word 'action-type 'deletes-forward) -(put 'delete-char 'action-type 'deletes-forward) -(put 'kill-sexp 'action-type 'deletes-forward) -(put 'backward-kill-sexp 'action-type 'deletes-backward) -(put 'backward-delete-char 'action-type 'deletes-backward) -(put 'delete-backward-char 'action-type 'deletes-backward) -(put 'backward-kill-word 'action-type 'deletes-backward) -(put 'backward-delete-char-untabify 'action-type 'deletes-backward) -(put 'slime-repl-newline-and-indent 'action-type 'inserts) -(put 'kill-ring-save 'action-type 'copies) - (defvar slime-presentation-map (make-sparse-keymap)) (define-key slime-presentation-map [mouse-2] 'slime-copy-presentation-at-point) @@ -2887,19 +2958,15 @@ end end." (let ((start (point))) (unless (bolp) (insert "\n")) (unless (string= "" result) - (slime-propertize-region `(face slime-repl-result-face) - (slime-propertize-region - (and slime-repl-enable-presentations - `(face slime-repl-result-face - slime-repl-old-output ,(- slime-current-output-id) - mouse-face slime-repl-output-mouseover-face - keymap ,slime-presentation-map)) - (insert result))) + (if slime-repl-enable-presentations + (slime-insert-presentation result slime-current-output-id) + (slime-propertize-region `(face slime-repl-result-face) + (insert (substring result 1)))) (unless (bolp) (insert "\n")) (let ((inhibit-read-only t)) (put-text-property (- (point) 2) (point) 'rear-nonsticky - '(slime-repl-old-output face read-only)))) + '(slime-repl-old-output slime-repl-presentation face read-only)))) (let ((prompt-start (point)) (prompt (format "%s> " (slime-lisp-package-prompt-string)))) (slime-propertize-region -- Matthias Koeppe -- http://www.math.uni-magdeburg.de/~mkoeppe

1 0

[slime-devel] feture suggjestion
by Knut Olav Bøhmer 23 Jun '05

23 Jun '05

When I'm in a buffer editing my lisp-code, and decide to test the code, by pressing C-x C-e, slime forks of a new thread with the name "worker". The problem starts when I have several threads running from different buffers with the name "worker", and one of the threads ends in an infinite loop or deadlock, and I want to kill one of the threads. Then I don't know which thread is doing what. So If the thread got the name "worker + buffer name" then I would have slightly more control. Good Idea? -- Knut Olav Bøhmer tlf: +47 47400777

1 0

[slime-devel] SLIME minimal hardware requirements
by borisman＠netmail.kg 22 Jun '05

22 Jun '05

My name is Boris, and i have a question that google can not answer. I want to buy a very old notebook and i wonder will i be able to run slime on it. The notebook is Toshiba Saellite 2100CS with i486 DX 25 processor and 8 Mb ram. I want to install Dewbian Linux on it and work in text mode. (no XWindows). Will i be able use EMACS+SLIME+LISP on it ??? Sorry bothering you, but don't know anybody else to ask.

4 4

[slime-devel] Daily ChangeLog diff
by lgorrie＠common-lisp.net 21 Jun '05

21 Jun '05

Index: slime/ChangeLog diff -u slime/ChangeLog:1.715 slime/ChangeLog:1.716 --- slime/ChangeLog:1.715 Mon Jun 13 11:34:53 2005 +++ slime/ChangeLog Tue Jun 21 20:28:58 2005 @@ -1,3 +1,7 @@ +2005-06-21 Edi Weitz <edi(a)agharta.de> + + * swank.lisp (find-matching-packages): Also use nicknames. + 2005-06-13 Edi Weitz <edi(a)agharta.de> * swank.lisp (list-all-systems-in-central-registry): Delete

1 0

[slime-devel] REPL buffer transcripts [PATCH]
by Richard M Kreuter 18 Jun '05

18 Jun '05

Hello, I fairly often copy text from the REPL buffer into text files and emails, and then spend a lot of time adding text to the beginnings of lines to indicate which text was input, output, error messages, or return values. But with a little tweaking, slime can do this stuff for me. I've attached a patch against today's anonymous cvs, with which I've made the following transcript. If people think this feature is worth including but that the code needs changes, please let me know what has to be done. -- Richard ; SLIME 2005-06-13 CL-USER> 1 1 CL-USER> ,transcript ; REPL transcript printing activated CL-USER> 1 => 1 CL-USER> (read) <- "asdf" => "asdf" CL-USER> (error "Foo.") !! Foo. !! [Condition of type SIMPLE-ERROR] ; Evaluation aborted CL-USER> (loop for input = (read) until (not (numberp input)) do (format t "~A~%~%~A~%" input (sqrt input)) finally (return input)) <- 1 -| 1 -| -| 1.0 <- 4 -| 4 -| -| 2.0 <- q => Q CL-USER> (values 5 "foo") => 5 => "foo" CL-USER> (macroexpand '(with-open-file (*standard-output* "foo") (read-line))) => (LET ((*STANDARD-OUTPUT* (OPEN "foo")) (#:G3835 T)) (UNWIND-PROTECT (MULTIPLE-VALUE-PROG1 (PROGN (READ-LINE)) (SETQ #:G3835 NIL)) (WHEN *STANDARD-OUTPUT* (CLOSE *STANDARD-OUTPUT* :ABORT #:G3835)))) => T CL-USER> ,transcript ; REPL transcript printing deactivated CL-USER> 1 1 CL-USER> (read) "asdf" "asdf" CL-USER> (error "Foo.") ; Evaluation aborted CL-USER>

1 0

[slime-devel] README clarification
by Brandon J. Van Every 16 Jun '05

16 Jun '05

The README does not match the .PDF documentation when explaining how to configure the ~/.emacs file. The README mentions setting the inferior-lisp-program variable, but it doesn't show how to do it. An example would be moderately helpful to people who are not familiar with emacs configuration, i.e. it may save someone a half hour of futzing. Believe it or not, people (like me :-) do things like wonder "what is this variable they're talking about?" and then start looking for it in the emacs docs, which of course it isn't in. Something like the following would be an improvement over what is currently written: (setq inferior-lisp-program "D:/lang/clisp-2.33.2/clisp.bat") (add-to-list 'load-path "D:/app/slime") (require 'slime) (slime-setup) Also the README doesn't say you need to put this stuff in your ~/.emacs file. Maybe there are other ways to get it working, but it might be good if the README assumes less familiarity with emacs rather than more. Generally, the .PDF documentation is clearer than the README. Maybe it would be best to synchronize the README from the .PDF doc. Cheers, www.indiegamedesign.com Brandon Van Every Seattle, WA "The pioneer is the one with the arrows in his back." - anonymous entrepreneur

1 0