August 2008 - slime-devel - mailman3.common-lisp.net

[slime-devel] [RfC] Per event bindings
by Tobias C. Rittweiler 20 Aug '08

20 Aug '08

Hi, I recently introduced Swank-Compilation-Units; within one such units the compiler notes of several compilation requests are accumulated. These units were introduced to implement C-c C-k in Xrefs buffers where several toplevel forms should be recompiled at once (the reason we don't do it with one request is the preservation of source-locations.) The current implementation works perfectly fine with threads. But there's a use case that will break for the serve-event communication style: Let's say you have some form which will activate some application's event loop, and you use C-c C-c to "evaluate" the form. This C-c C-c will bind *SWANK-COMPILATION-UNIT*. All consequent C-c C-c's will be executed in a dynamic environment where *SWANK-COMPILATION-UNIT* is still bound to that the unit of the application-event-loop evaluation. This means that the compilation of all consequent C-c C-c's will be accumulated, which means they won't be displayed at all (as long as the event loop is running.) The reason behind this behaviour is that I do (defvar *swank-compilation-unit* nil) (defmacro with-swank-compilation-unit (...) ... `(let ((*swank-compilation-unit* (or *swank-compilation-unit* (make-swank-compilation-unit)))) ...)) which works perfectly reasonable for threads, but not for SERVE-EVENT, as that will interrupt, and execute new code within the dynamic contour of the interrupted code. A solution to this is to introduce per event bindings which are implemented differently depending on the communication-style used (for :spawn, thread-local bindings can be used; for :fd-handler, we'd need some mapping from some request-id to the value.) Opinions? -T. PS: If the form starting the application's event loop is started by using C-M-x rather than C-c C-c this problem doesn't occur, as C-M-x doesn't involve Swank-Compilation-Units. It could be argued that C-M-x is the right keybinding to evaluate the form in the first place. So there's the question if the abovely sketched solution is worthwhile at all. Of course, we may be bitten by this in future in some form which manifests itself in a much more obscure way than the bug described above. This is a kinda handwaving argument, but I wanted to raise the issue anyway.

2 2

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 20 Aug '08

20 Aug '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1449 slime/ChangeLog:1.1453 --- slime/ChangeLog:1.1449 Tue Aug 12 13:54:43 2008 +++ slime/ChangeLog Thu Aug 14 11:32:40 2008 @@ -1,3 +1,47 @@ +2008-08-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + If another frame is already displaying a REPL, `C-c C-z' will now + select this window instead of splitting the current frame, and + displaying the REPL in the newly created window. + + * slime.el (slime-switch-to-output-buffer-search-all-frames): New + variable to customize this behaviour, as it may not be optimal for + people using multiple screens at once. + (slime-switch-to-output-buffer): Adapted accordingly. + +2008-08-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + Xref buffers: `M-,' in an Xref buffer will now revert to an Emacs + state as of before the Xref command. + + * slime.el (slime-xref-mode-map): Add `slime-xref-retract' as `M-,' + (slime-xref-retract): New. Restores the emacs snapshot unconditionally. + (slime-xref-quit): Use it. + +2008-08-14 Tobias C. Rittweiler <tcr(a)freebits.de> + + Xref buffers: `q', and `SPC' will push onto the find-definition + stack such that M-, will work afterwards. + + * slime.el (defstruct slime-emacs-snapshot): Also save the point + explicitly. It is implicitly stored already in the + window-configuration, but inaccessible therein. + (slime-current-emacs-snapshot, slime-set-emacs-snapshot): Adapted. + (slime-push-definition-stack-from-snapshot): New. Reason for above + changes. + + (slime-with-popup-buffer): Make sure that the current + emacs-snapshot is taken, not only when the :emacs-snapshot + argument is NIL at expansion time, but also on runtime. + + (slime-with-xref-buffer): The name of the Xref buffer was created + at expansion time, but must be computed at runtime. Fix that. + (slime-xref-quit): Use `slime-push-definition-stack-from-snapshot' + (sime-xref-goto): Adapted to also push onto the stack. + + * slime.el (slime-compute-modeline-package): Cleaned up. + (slime-update-modeline-string): Ditto. + 2008-08-12 Helmut Eller <heller(a)common-lisp.net> Add a dump-image function to the loader.

4 10

[slime-devel] Unicode support for CMUCL
by Raymond Toy 19 Aug '08

19 Aug '08

Here is a small patch to enable support for Unicode for CMUCL. Seems to work. May need some more work on the *external-format-to-coding-system*, but iso-8859-1 and utf-8 appear to work ok, for the little bit of testing I've done. Ray --- swank-cmucl.lisp-orig 2008-06-01 18:06:07.000000000 -0400 +++ swank-cmucl.lisp 2008-06-01 18:01:44.000000000 -0400 @@ -99,10 +99,13 @@ (ext:close-socket fd))) (defimplementation accept-connection (socket &key - external-format buffering timeout) - (declare (ignore timeout external-format)) + external-format buffering timeout) + (declare (ignore timeout)) (let ((buffering (or buffering :full))) - (make-socket-io-stream (ext:accept-tcp-connection socket) buffering))) + (make-socket-io-stream (ext:accept-tcp-connection socket) + buffering + :external-format + (or external-format :iso-8859-1)))) ;;;;; Sockets @@ -117,10 +120,23 @@ (let ((hostent (ext:lookup-host-entry hostname))) (car (ext:host-entry-addr-list hostent)))) -(defun make-socket-io-stream (fd buffering) +(defvar *external-format-to-coding-system* + '((:iso-8859-1 + "latin-1" "latin-1-unix" "iso-latin-1-unix" + "iso-8859-1" "iso-8859-1-unix") + (:utf-8 "utf-8" "utf-8-unix"))) + +(defimplementation find-external-format (coding-system) + (car (rassoc-if (lambda (x) (member coding-system x :test #'equal)) + *external-format-to-coding-system*))) + + +(defun make-socket-io-stream (fd buffering &key external-format) "Create a new input/output fd-stream for FD." (sys:make-fd-stream fd :input t :output t :element-type 'base-char - :buffering buffering)) + :buffering buffering + #+unicode :external-format + #+unicode external-format)) ;;;;; Signal-driven I/O

3 7

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 19 Aug '08

19 Aug '08

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.120 slime/contrib/ChangeLog:1.121 --- slime/contrib/ChangeLog:1.120 Tue Aug 12 13:54:37 2008 +++ slime/contrib/ChangeLog Mon Aug 18 05:20:21 2008 @@ -1,3 +1,8 @@ +2008-08-18 Ariel Badichi <abadichi(a)bezeqint.net> + + * slime-fuzzy.el: Rename `add-local-hook' to `slime-add-local-hook'. + * slime-presentations.el: Ditto. + 2008-08-12 Helmut Eller <heller(a)common-lisp.net> * slime-clipboard.el (slime-clipboard-insert-ref): Set

1 0

[slime-devel] Missed `add-local-hook' renames
by Ariel Badichi 18 Aug '08

18 Aug '08

Hello, There are a couple of missing renames for `add-local-hook', which prevent SLIME from loading properly in Emacs if either `slime-fuzzy' or `slime-presentation' contribs are used. The following patch fixes that. Ariel commit feb5c8e7ba42ef14708ab2c1304e8d7cea8bb5cb Author: Ariel Badichi <abadichi(a)bezeqint.net> Date: Mon Aug 18 04:38:44 2008 +0300 More `add-local-hook' to `slime-add-local-hook' renames. diff --git a/contrib/slime-fuzzy.el b/contrib/slime-fuzzy.el index 42fc546..bbae454 100644 --- a/contrib/slime-fuzzy.el +++ b/contrib/slime-fuzzy.el @@ -359,7 +359,7 @@ done." (when (boundp 'window-configuration-change-hook) (add-hook 'window-configuration-change-hook 'slime-fuzzy-window-configuration-change)) - (add-local-hook 'kill-buffer-hook 'slime-fuzzy-abort) + (slime-add-local-hook 'kill-buffer-hook 'slime-fuzzy-abort) (setq buffer-quit-function 'slime-fuzzy-abort)) ; M-Esc Esc (when slime-fuzzy-completion-in-place ;; switch back to the original buffer diff --git a/contrib/slime-presentations.el b/contrib/slime-presentations.el index 3d298db..6da488b 100644 --- a/contrib/slime-presentations.el +++ b/contrib/slime-presentations.el @@ -858,8 +858,8 @@ even on Common Lisp implementations without weak hash tables." (lambda () ;; Respect the syntax text properties of presentation. (set (make-local-variable 'parse-sexp-lookup-properties) t) - (add-local-hook 'after-change-functions - 'slime-after-change-function))) + (slime-add-local-hook 'after-change-functions + 'slime-after-change-function))) (add-hook 'slime-event-hooks 'slime-dispatch-presentation-event) (setq slime-write-string-function 'slime-presentation-write) (add-hook 'slime-repl-return-hooks 'slime-presentation-on-return-pressed)

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 18 Aug '08

18 Aug '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1456 slime/ChangeLog:1.1465 --- slime/ChangeLog:1.1456 Fri Aug 15 18:59:10 2008 +++ slime/ChangeLog Sun Aug 17 19:01:18 2008 @@ -1,3 +1,56 @@ +2008-08-18 Helmut Eller <heller(a)common-lisp.net> + + * swank.lisp (install-fd-handler): Bind *emacs-connection* with + with-connection, for case when the signal hander is called out of + the blue. + (swank-debugger-hook): Don't assume that the hook argument is + #'swank-debugger-hook itself. + + * test.sh (usage): Use cat rather not echo for here-documents. + +2008-08-17 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-inspector-show-source): New command. + (slime-inspector-mode-map): Bind it to ".". + (sldb-highlight-sexp): Use slime-flash-region rather than a + permanent overlay. + (slime-popup-buffer-quit): Must call bury-buffer without argument. + + * swank.lisp (find-source-location-for-emacs): New function. + + * slime.el (slime-add-local-hook): Renamed from add-local-hook. + (slime-switch-to-output-buffer): Drop the connection argument. It + was never used. + (slime-switch-to-output-buffer-search-all-frames): Deleted. Use + display-buffer-reuse-frames instead. + (slime-switch-to-output-buffer): Use pop-to-buffer to select the + window and frame. + +2008-08-15 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-switch-to-output-buffer): Fix regression + discovered by Ariel Badichi. Programmatic invocation of this + function expect this function to also set-buffer to the REPL + buffer. Even though that's ugly, we do it for now, and declare it + to be a FIXME. + (slime-dispatch-event): Don't use `message' for the pipelined + request message, but `slime-display-oneliner' which will truncate + the form to be displayed. + +2008-08-17 Helmut Eller <heller(a)common-lisp.net> + + * swank-cmucl.lisp (waitpid): Don't use unix::pid-t, it's only + defined for Linux. + + * swank-sbcl.lisp (sb-thread::get-foreground): Don't override. + Let SBCL people fix this. + +2008-08-17 B.Scott Michel <scooter.phd(a)gmail.com> + + * swank-loader.lisp (dump-image): Move this function to the end of + the file to stop SBCL from reporting two times the same false + alarm. + 2008-08-15 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-popup-buffer-quit): Fix regression; we have to

1 0

[slime-devel] New event handling
by Tobias C. Rittweiler 17 Aug '08

17 Aug '08

Hi Helmut, Your recent polishing of the event handling related code is truly appreciated. Perhaps you're already aware of the following issues, but just haven't come around dealing with them. For the case they're news to you: On SBCL, a) when using :spawn, a lot of tests are failing because of timeout waiting. b) when using :fd-handler, interrupting via C-c C-c results in a Type-Error. c) when toggling *log-events* at the REPL, some warning is signalled. It seems to work all fine, though. -T.

2 1

[slime-devel] READ broken in latest SLIME
by Ariel Badichi 17 Aug '08

17 Aug '08

Hello, If I enter `(read)' at the SLIME REPL, Emacs gives me the following messages: error in process filter: slime-mark-output-end: Wrong type argument: integer-or-marker-p, nil error in process filter: Wrong type argument: integer-or-marker-p, nil The offending change seems to be tcr's `C-c C-z' patch from 2008-08-14: merely selecting the REPL window is apparently not enough. After setting the current buffer to be the REPL buffer, READ works as expected. Ariel

2 1

[slime-devel] Slime inspector + large arrays
by Bart Botta 17 Aug '08

17 Aug '08

With current CVS, the slime inspector shows the entire contents of large arrays when they are in slots of a class instance being inspected. for example: CL-USER> (defclass a () ((b :initform (make-array 10000 :initial-contents (loop for i below 10000 collect i))))) (make-instance 'a) #<A {23F76299}> when inspecting the object, all of the array is displayed, but when you look at just the slot, it is limited to a more reasonable 120 or so entries. -b-

1 0

[slime-devel] Breakage on cmucl 19e
by Harald Hanche-Olsen 17 Aug '08

17 Aug '08

;; Error while loading: home:.slime/fasl/2008-08-15/cmu-19e_(19e)-unix-x86/swank-cmucl.x86f ;; Condition: (during macroexpansion) ;; ;; Error in function alien::%parse-alien-type: Unknown alien type: unix::pid-t ;; Aborting. The reference to unix::pid-t was introduced with the latest edit (revision 1.187, adding save-image) of swank-cmucl.lisp. - Harald

2 1