August 2007 - slime-devel - mailman3.common-lisp.net

[slime-devel] Can you afford to lose 300, 000 potential customers per year ?
by Instant＠common-lisp.net 29 Aug '07

29 Aug '07

1 0

[slime-devel] Bug in slime-repl-insert-prompt
by Austin Haas 29 Aug '07

29 Aug '07

The most recent change in CVS: 2007-08-28 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> * slime.el (slime-repl-insert-prompt): Don't go to point-max but to slime-repl-input-start-mark if there is one. This fixes user input type-ahead again (this change from 2007-08-25 got lost). Testcase: Type (dotimes (i 5) (format t "Number ~A~%" i) (sleep 1)) and then type ahead while the command is executing and output arrives. causes the output from one expr in the repl to be placed AFTER the next repl prompt. e.g., CL-USER> (+ 1 2) CL-USER> 3 Desired result: CL-USER> (+ 1 2) 3 CL-USER> -austin

2 1

[slime-devel] repl output behavior
by Austin Haas 28 Aug '07

28 Aug '07

Hi, I just upgraded to the most recent cvs slime this afternoon (previous version was probably about a month old, but I'm not positive), and I noticed that the output in the repl shows up in a different way than I've ever seen. Now, any expression I evaluate (C-x C-e) in a buffer is echoed in the repl in a bottom to top order, starting from above the command prompt. Any output, warnings, etc., appear top to bottom above the echoed expression. Previously, everything flowed from top to bottom with expressions and output intertwined. Was the repl buffer's behavior changed or am I seeing something unintentional? Is there some way to get it back the way that it used to be? -austin

2 3

[slime-devel] OK to move arbitrary :write-string TARGET handler into core?
by Matthias Koeppe 28 Aug '07

28 Aug '07

Helmut, can I move the code to handle arbitrary :write-string TARGETs (via a hashtable of markers) from slime-presentations into core SLIME (see diff)? This code is actually not related to presentations, but is for separating out streams like *trace-output*. Matthias --- slime.el 28 Aug 2007 08:52:04 -0700 1.832 +++ slime.el 28 Aug 2007 09:34:05 -0700 @@ -2970,7 +2970,52 @@ (if (>= (marker-position slime-output-end) (point)) ;; If the output-end marker was moved by our insertion, ;; set it back to the beginning of the REPL result. - (set-marker slime-output-end result-start))))))) + (set-marker slime-output-end result-start))))) + (t + (let* ((marker (slime-output-target-marker target)) + (buffer (and marker (marker-buffer marker)))) + (when buffer + (with-current-buffer buffer + (save-excursion + ;; Insert STRING at MARKER, then move MARKER behind + ;; the insertion. + (goto-char marker) + (insert-before-markers string) + (set-marker marker (point))))))))) + +(defvar slime-last-output-target-id 0 + "The last integer we used as a TARGET id.") + +(defvar slime-output-target-to-marker + (make-hash-table) + "Map from TARGET ids to Emacs markers that indicate where +output should be inserted.") + +(defun slime-output-target-marker (target) + "Return a marker that indicates where output for TARGET should +be inserted." + (case target + ((nil) + (with-current-buffer (slime-output-buffer) + slime-output-end)) + (:repl-result + (with-current-buffer (slime-output-buffer) + slime-repl-input-start-mark)) + (t + (gethash target slime-output-target-to-marker)))) + + +(defun slime-redirect-trace-output () + "Redirect the trace output to a separate Emacs buffer." + (interactive) + (let ((buffer (get-buffer-create "*SLIME Trace Output*"))) + (with-current-buffer buffer + (let ((marker (copy-marker (buffer-size))) + (target (incf slime-last-output-target-id))) + (puthash target marker slime-output-target-to-marker) + (slime-eval `(swank:redirect-trace-output ,target)))) + (pop-to-buffer buffer))) + (defun slime-switch-to-output-buffer (&optional connection) "Select the output buffer, preferably in a different window." -- Matthias Köppe -- http://www.math.uni-magdeburg.de/~mkoeppe (currently @math.ucdavis.edu)

2 1

[slime-devel] OK to simplify :write-string message incompatibly?
by Matthias Koeppe 28 Aug '07

28 Aug '07

Helmut, I would like to simplify the :write-string message from (:write-string output &optional id target) to (:write-string output &optional target). (The reason is that ID is not used in core SLIME anymore; and I also do not need it for supporting presentations.) OK to change, or do we need to care about compatibility of the SLIME protocol at this point? Matthias -- Matthias Köppe -- http://www.math.uni-magdeburg.de/~mkoeppe (currently @math.ucdavis.edu)

2 2

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 28 Aug '07

28 Aug '07

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1171 slime/ChangeLog:1.1176 --- slime/ChangeLog:1.1171 Sun Aug 26 19:36:43 2007 +++ slime/ChangeLog Mon Aug 27 11:48:58 2007 @@ -1,3 +1,47 @@ +2007-08-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-make-extended-operator-parser/look-ahead): Move + to end of symbol at point. + (slime-make-form-spec-from-string): Fixes unexpected behaviour of + `save-excursion'. + +2007-08-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-sexp-at-point): Fixes a few edge cases were + Emacs' `(thing-at-point 'sexp)' behaves suboptimally. For example, + `foo(bar baz)' where point is at the ?\(. + (slime-internal-scratch-buffer): New. This variable holds an + internal scratch buffer that can be reused instead of having to + create a new temporary buffer again and again. + (slime-make-extended-operator-parser/look-ahead): Uses + `slime-make-form-spec-from-string' to parse nested expressions + properly. + (slime-nesting-until-point): Added docstring. + (slime-make-form-spec-from-string): Added new optional parameter + for stripping the operator off the passed string representation of + a form. Necessary to work in the context of + `slime-make-extended-operator-parser/look-ahead'. Added safety check + against a possible endless recursion. + + * swank.lisp (parse-form-spec): Looses restriction for nesting. + +2007-08-27 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-eval-feature-conditional): Fix typo. + (slime-keywordify): Simplify. + +2007-08-27 Helmut Eller <heller(a)common-lisp.net> + + Move presentations to contrib. Part I. + + * slime.el (slime-event-hooks, slime-dispatch-event): New hook. + (slime-write-string-function, slime-write-string): New hook. + (slime-repl-return-hooks, slime-repl-return): New hook. + (slime-repl-current-input-hooks, slime-repl-current-input): New hook. + (slime-open-stream-hooks, slime-open-stream-to-lisp): New hook. + (sldb-insert-locals, slime-inspector-insert-ispec) + (slime-last-expression): Don't use presentations. + 2007-08-26 Tobias C. Rittweiler <tcr(a)freebits.de> Reduces needless interning of symbols that was introduced by my

1 0

[slime-devel] SWANK and *read-eval*
by Tobias C. Rittweiler 28 Aug '07

28 Aug '07

Hi, *read-eval* is currently set to T everywhere something is READ. While I don't currently see anything that could be (at least easily) exploitable, it still leaves an uneasy feeling in my stomach. Especially in the present of things like autodoc. Therefore I propose the following: (i) Setting *READ-EVAL* to NIL in CALL-WITH-BUFFER-SYNTAX. (ii) Introducing a new macro WITH-SAVE-IO-SYNTAX that expands to (WITH-STANDARD-IO-SYNTAX (LET ((*READ-EVAL* NIL)) ...)) and replacing all occurences of WITH-STANDARD-IO-SYNTAX with this new macro. Any thoughts on this issue? -T.

3 8

[slime-devel] slime-ppcre
by Kilian Sprotte 27 Aug '07

27 Aug '07

Hi, I would like to show you my code that I wrote for grepping through lisp code and for replacing stuff using cl-ppcre. Personally, I find this quite useful - I am curious of course, what you think. You find some informations here: http://www.cliki.net/SLIME-PPCRE There is one thing to note for getting this to load: I had to change swank-loader.lisp: (defvar *contribs* '(swank-fuzzy swank-ppcre) "List of names for contrib modules.") I was doing this on OSX, so I hope it works at least over here... :) Cheers, Kilian

2 1

[slime-devel] [patch] a way to specify custom Emacs indentation right in the Lisp source
by Michael Livshin 27 Aug '07

27 Aug '07

hi all, I find that sometimes the usual Slime way of determining macro indentation based on the position of the &BODY parameter in the argument list is not quite enough. for example: if the body of a certain macro is really a list of function bindings, then one would like those bindings to be indented properly. as it happens, Emacs (or rather common-lisp-indent-function) can be told exactly how to do that, but this information cannot be expressed by the macro argument list alone. the attached patch adds a way to define and access custom indentation specifications like this: <example> (defmacro my-macro (name (&rest args) &body function-bindings) ...) #+swank (setf (swank:emacs-indentation 'my-macro) '(6 4 (&whole 2 &rest (&whole 1 &lambda &body)))) </example> [ which won't work exactly as written, of course, because unfortunately Swank does not push anything onto *FEATURES*... ] thanks, --m * swank.lisp (*emacs-indentation*): New hash table, used to hold custom indentation specs. (emacs-indentation): Accessor for the above. (symbol-indentation): Use it. * slime.el (slime-clean-indent-spec): New function, removes CL package prefixes from symbols sent from Lisp. (slime-handle-indentation-update): Use it.

6 11

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 27 Aug '07

27 Aug '07

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1165 slime/ChangeLog:1.1171 --- slime/ChangeLog:1.1165 Sat Aug 25 16:05:00 2007 +++ slime/ChangeLog Sun Aug 26 19:36:43 2007 @@ -1,3 +1,83 @@ +2007-08-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + Reduces needless interning of symbols that was introduced by my + recent work on autodoc to a minimum. Also fixes this issue for + `slime-complete-form' which always interned symbols even before my + changes. + + * slime.el (slime-sexp-at-point): If N is given, but there aren't + N sexps available at point, make it return a list of just as many + as there are. + (slime-make-form-spec-from-string): New. Creates a ``raw form + spec'' from a string that's suited for determining newly interned + symbols later in Swank. + (slime-parse-extended-operator/declare): Uses it. + + * swank.lisp (parse-symbol): Returns internal knowledge, to + provide a means for callers to perform a sanity check. + (call-with-ignored-reader-errors): New. Abstracted out from + `read-incomplete-form-from-string.' + + * swank.lisp (read-form-spec): New. Only READs elements of a form + spec if necessary. And if it does have to READ, it keeps track + of newly interned symbols which are returned as secondary + return value. + (parse-form-spec): Use it. Propagate newly interned symbols. + (parse-first-valid-form-spec): Likewise. + (arglist-for-echo-area, complete-form, completions-for-keyword): + Adapted to unintern the newly interned symbols. + + +2007-08-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (current-slime-narrowing-configuration): + Renamed to `slime-current-narrowing-configuration'. + (set-slime-narrowing-configuration): + Renamed to `slime-set-narrowing-configuration'. + (current-slime-emacs-snapshot): + Renamed to `slime-current-emacs-snapshot'. + (current-slime-emacs-snapshot-fingerprint): + Renamed to `slime-current-emacs-snapshot-fingerprint'. + (set-slime-emacs-snapshot): + Renamed to `slime-set-emacs-snapshot'. + +2007-08-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (save-restriction-if-possible): Fixed another typo, + duh! Thanks again to Matthias Koeppe. + +2007-08-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-cl-symbol-name): Handle vertical bars (|) + (%slime-nesting-until-point): Renamed to `slime-nesting-until-point'. + +2007-08-25 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + Fix a bug where REPL results would sometimes be indented by a + random amount. + + * slime.el (slime-insert-presentation): Make the + rectangle-ification of multi-line presentations, introduced + 2006-12-19, optional. + (slime-write-string): Use it here only for regular output, but not + for REPL results. + (sldb-insert-locals): Use it here. + (slime-inspector-insert-ispec): Use it here. + +2007-08-25 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> + + Fix handling of user-input type-ahead in the REPL. + Reported by Madhu <enometh(a)meer.net> on 2007-04-24. + + * slime.el (slime-write-string): Make sure text properties are + rear-nonsticky, so typed-ahead user input does not pick up the + text properties. Fix up some markers. + (slime-reset-repl-markers): Make the marker slime-output-end of + insertion type nil (no automatic advances on insertions). + (slime-with-output-end-mark): Update the location of + slime-output-end here manually. + (slime-repl-update-banner): Use insert-before-markers. + 2007-08-25 Matthias Koeppe <mkoeppe(a)mail.math.uni-magdeburg.de> New command slime-redirect-trace-output creates a separate Emacs @@ -5,14 +85,16 @@ * slime.el (slime-last-output-target-id): New variable. (slime-output-target-to-marker): New variable. + (slime-output-target-marker): New function. (slime-write-string): Handle general "target" arguments using - slime-output-target-to-marker. + slime-output-target-marker. (slime-redirect-trace-output): New command. (slime-easy-menu): Add a menu item for it. * slime.el (slime-mark-presentation-start) (slime-mark-presentation-end): Make "target" argument optional. - + Use slime-output-target-to-marker. + * swank.lisp (make-output-stream-for-target): New function, factored out from open-streams. (open-streams): Use it here.

1 0