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

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

20 Jul '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1371 slime/ChangeLog:1.1372 --- slime/ChangeLog:1.1371 Thu Jul 17 18:49:36 2008 +++ slime/ChangeLog Sat Jul 19 07:34:19 2008 @@ -1,3 +1,21 @@ +2008-07-19 Tobias C. Rittweiler <tcr(a)freebits.de> + + REPL shortcuts now leave an appropriate Common Lisp form in the + REPL history. + + * slime.el (slime-within-repl-shortcut-handler-p): New global. T + if truly inside a repl shortcut handler invoked by ,foo on the + REPL. + (slime-handle-repl-shortcut): Bind above global appropriatly. + (slime-repl-shortcut-eval): New; should be used in repl shortcut + handlers instead of `slime-eval'. + (slime-repl-shortcut-eval-async): New; should be used in repl + shortcut handlers instead of `slime-eval-async'. + (defslime-repl-shortcut): Update docstring. + (slime-repl-set-package): Use slime-repl-shortcut-eval. + (slime-set-default-directory): Ditto. + (slime-sync-package-and-default-directory): Ditto. + 2008-07-18 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-recompile-locations): Locations were potentially Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.108 slime/contrib/ChangeLog:1.109 --- slime/contrib/ChangeLog:1.108 Wed Jul 16 14:44:28 2008 +++ slime/contrib/ChangeLog Sat Jul 19 07:39:23 2008 @@ -1,3 +1,7 @@ +2008-07-19 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-asdf.el (slime-oos): Use `slime-repl-shortcut-async'. + 2008-07-16 Tobias C. Rittweiler <tcr(a)freebits.de> * swank-asdf.lisp (operate-on-system-for-emacs): Wrapped in

1 0

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

18 Jul '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1369 slime/ChangeLog:1.1371 --- slime/ChangeLog:1.1369 Wed Jul 16 15:18:51 2008 +++ slime/ChangeLog Thu Jul 17 18:49:36 2008 @@ -1,3 +1,28 @@ +2008-07-18 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-recompile-locations): Locations were potentially + recompiled within a wrong package. Fix that. + +2008-07-18 Tobias C. Rittweiler <tcr(a)freebits.de> + + An explicit numeric value as prefix-arg given to `C-c C-c' will + now represent the debug level the defun is compiled with; + `C-u C-c C-c' defaults to maximum debug like before. (Now also + works for recompilation commands in xref buffers.) + + * slime.el (slime-compilation-debug-level): Renamed from + `slime-compile-with-maximum-debug'. + (slime-normalize-optimization-level): New. + (slime-compile-defun): Adapted accordingly. + (slime-compile-region): Ditto. + (slime-recompile-location): Added setting of debug-level. + (slime-recompile-locations): Ditto. + (slime-recompile-xref): Now takes debug-level prefix-arg. + (slime-recompile-all-xrefs): Ditto. + + * swank-sbcl.lisp (defimplementation swank-compile-string): + Adapted accordingly. + 2008-07-16 Tobias C. Rittweiler <tcr(a)freebits.de> * slime.el (slime-xref-dspec-at-point): Make more robust.

1 0

[slime-devel] Unicode characters
by Stas Boukarev 17 Jul '08

17 Jul '08

When I enter in the slime's repl (string #\u4C36), Slime enters debugger and REPL is no more usable: Debugger entered: nil (condition-case error (slime-net-read) (error (debug) (slime-net-close process t) (error "net-read error: %S" error))) slime-net-read-or-lose(#<process SLIME Lisp>) slime-process-available-input(#<process SLIME Lisp>) slime-net-filter(#<process SLIME Lisp> #("000024(:presentation-start 2 :repl-result)000024(:write-string \"\\\"\xe4\xb0\xb6\\\"\" :repl-result)000022(:presentation-end 2 :repl-result)000020(:write-string \"\n\" :repl-result)000016(:return (:ok nil) 17)" 0 66 nil 66 69 (composition (...) help-echo utf-8-help-echo untranslated-utf-8 19510) 69 192 nil)) slime-net-coding-system is 'utf-8-unix, Slime is from cvs, GNU Emacs 22.2.1 -- With Best Regards, Stas.

2 1

[slime-devel] Bug in CLEAR-INPUT
by Stelian Ionescu 17 Jul '08

17 Jul '08

If I run the example in the CLHS page of CLEAR-INPUT, Example: HU> (defun read-sleepily (&optional (clear-p nil) (zzz 0)) (list (progn (print '>) (read)) ;; Note that input typed within the first ZZZ seconds ;; will be discarded. (progn (print '>) (if zzz (sleep zzz)) (print '>>) (if clear-p (clear-input)) (read)))) READ-SLEEPILY HU> (read-sleepily t 10) > 10 > 20 30 >> End Example I get this: Recursive lock attempt #S(SB-THREAD:MUTEX :NAME "buffer read lock" :%OWNER #<SB-THREAD:THREAD "repl-thread" RUNNING {BC10BB1}> :STATE 1). [Condition of type SIMPLE-ERROR] Backtrace: 0: (SB-THREAD:GET-MUTEX #<unavailable argument> #<unavailable argument> #<unavailable argument>) 1: ((FLET #:WITHOUT-INTERRUPTS-BODY-[CALL-WITH-MUTEX]477)) 2: ((LAMBDA (SWANK-BACKEND::LOCK FUNCTION)) #S(SB-THREAD:MUTEX :NAME "buffer read lock" :%OWNER #<SB-THREAD:THREAD "repl-thread" RUNNING {BC10BB1}> :STATE 1) #<CLOSURE (LAMBDA NIL) {C1FDE25}>) 3: (SWANK-BACKEND:CALL-WITH-LOCK-HELD #S(SB-THREAD:MUTEX :NAME "buffer read lock" :%OWNER #<SB-THREAD:THREAD "repl-thread" RUNNING {BC10BB1}> :STATE 1) #<CLOSURE (LAMBDA NIL) {C1FDE25}>) 4: ((SB-PCL::FAST-METHOD SB-GRAY:STREAM-CLEAR-INPUT (SWANK-BACKEND::SLIME-INPUT-STREAM)) #(3 NIL 4 NIL 5 NIL) #<unused argument> #<SWANK-BACKEND::SLIME-INPUT-STREAM {B9C4011}>) 5: (CLEAR-INPUT #<SWANK-BACKEND::SLIME-INPUT-STREAM {B9C4011}>) 6: (SWANK::CLEAR-USER-INPUT) 7: (SWANK::REPL-EVAL "30 ") -- Stelian Ionescu a.k.a. fe[nl]ix Quidquid latine dictum sit, altum videtur.

2 1

[slime-devel] Weird Slime Loading Behavior & Less Pretty Autodoc
by Yacin Nadji 17 Jul '08

17 Jul '08

I'm running CVS slime on Mac OS X, and I'm having some strange issues. I upgraded to CVS slime because I broke my version I had installed through macports by doing some silly upgrades of sbcl. First, SLIME-SETUP seems to load things improperly. If you take a look at my .emacs[1] you'll see I had to add (require 'slime-fuzzy) in order to get fuzzy auto-complete working. Looking at the source for what slime-fancy does, this should be auto-loaded, but it doesn't seem to work properly without manually loading it. I have a feeling this has to do with my other problem. Currently, the little function prototype helper[2] (that says the name of the function, and the name of the parameters it takes) looks a lot uglier than it used to. It used to have colors, and highlighted the current parameter I should be entering in. I realize this isn't a huge deal, but I wanted to see if it was due to using CVS slime, or if I just have something configured incorrectly. Thanks! Yacin [1] http://s3.getdropbox.com/u/10591/emacs.txt [2] http://s3.getdropbox.com/u/10591/slime-autodoc.png

2 3

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 17 Jul '08

17 Jul '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1367 slime/ChangeLog:1.1369 --- slime/ChangeLog:1.1367 Sat Jul 5 09:38:25 2008 +++ slime/ChangeLog Wed Jul 16 15:18:51 2008 @@ -1,3 +1,64 @@ +2008-07-16 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-xref-dspec-at-point): Make more robust. + (slime-xref-insert-recompilation-flags): Ditto. + (slime-column-max): New. + +2008-07-16 Tobias C. Rittweiler <tcr(a)freebits.de> + + Recompilation support added to xref buffers. You can now use + `C-c C-c' in an xref buffer to recompile the defun represented by + the xref at point. Similiarly, you can use `C-c C-k' to recompile + all xrefs displayed. + + For example, if you've changed a macro, and want to recompile all + the functions in the image which use that macro, you first call + `slime-who-macroexpands' (C-c C-w RET), and then issues `C-c C-k' + in the xref buffer that just popped up. + + [There's no guarantee that this will actually recompile all + functions that depend on the changed macro, as this obviously + depends on the quality of the backend's WHO-MACROEXPANDS + implementation.] + + * swank.lisp: Introduced the notion of a SWANK-COMPILATION-UNIT, + so we're able to compile different stuff comming from Slime one + after the other, and have compiler notes &c. collected in a + contiguous manner. + + (defstruct :swank-compilation-unit): New. Contains compilation + notes, compilation results, etc. + (*swank-compilation-unit*): New. Current Swank Compilation Unit. + (with-swank-compilation-unit): New. Like WITH-COMPILATION-UNIT. + (swank-compilation-unit-for-emacs): New. + (swank-compiler): Adapted; collect compilation stuff into the + current swank-compilation-unit. + (compile-string-for-emacs): Use WITH-SWANK-COMPILATION-UNIT. + (compile-file-for-emacs): Ditto. + + (*compiler-notes*, clear-compiler-notes): Removed. + (compiler-notes-for-emacs): Removed. + + * slime.el (slime-compilation-unit, slime-last-compilation-unit), + (slime-compiler-notes, slime-compiler-results): New/Adapted. + + (slime-make-compile-expression-for-swank): Factored out from + `slime-compile-string'. + (slime-recompile-location): New. + (slime-recompile-locations): New. + (slime-pop-to-location): &optional `where' arg can now also be + 'excursion to only reset the current-buffer, but not switch. + + (slime-xref-mode-map): Add `C-c C-c' and `C-c C-k'. + (slime-xref-dspec-at-point): New. + (slime-all-xrefs): New. + (slime-recompile-xref): New. + (slime-recompile-all-xrefs): New. + (slime-make-xref-recompilation-cont): New. + (slime-xref-inert-recompilation-flags): New. + + (slime-trim-whitespace): New utility. + 2008-07-05 Tobias C. Rittweiler <tcr(a)freebits.de> * swank.lisp: Revert Melis' change from 2008-07-04; Global IO Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.107 slime/contrib/ChangeLog:1.108 --- slime/contrib/ChangeLog:1.107 Sat Jun 7 07:46:15 2008 +++ slime/contrib/ChangeLog Wed Jul 16 14:44:28 2008 @@ -1,3 +1,8 @@ +2008-07-16 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-asdf.lisp (operate-on-system-for-emacs): Wrapped in + WITH-SWANK-COMPILATION-UNIT. + 2008-06-07 Tobias C. Rittweiler <tcr(a)freebits.de> * slime-parse.el (slime-cl-symbol-name),

1 0

[slime-devel] basic sldb question
by Stever 13 Jul '08

13 Jul '08

I am teaching myself lisp, but sometimes I like to step through other people's code using a debugger. I'm trying to figure out how the sldb debugger works in slime, but I am having trouble doing so. The documentation here: http://common-lisp.net/project/slime/doc/html/Examining-frames.html#Examini… talks about "frames". Ummm, what's a frame? How do I stop execution of code in slime/emacs and run each line one by one? I checked documentation and I searched around in google, but strangely enough I can't find much on it. - Steve

3 2

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 06 Jul '08

06 Jul '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1365 slime/ChangeLog:1.1367 --- slime/ChangeLog:1.1365 Fri Jul 4 19:30:44 2008 +++ slime/ChangeLog Sat Jul 5 09:38:25 2008 @@ -1,3 +1,31 @@ +2008-07-05 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp: Revert Melis' change from 2008-07-04; Global IO + redirection seems currently to be broken, and while it's not due + to that commit (it seems that it's been broken since longer), I + want to be on a safe bet. + +2008-07-05 Tobias C. Rittweiler <tcr(a)freebits.de> + + `M-x slime-lisp-threads' will now contain a summary of what's + currently executed in a thread that was created by Swank. + + * swank-backend.lisp (thread-description, set-thread-description): + New interface functions to associate strings with threads. + * swank-sbcl.lisp (thread-description, set-thread-description): + Implemented. + + * swank.lisp (call-with-thread-description), + (with-thread-description): New. + (read-from-emacs): Now temporarily sets the thread-description of + the current thread to a summary of what's going to be executed by + the current request. + (defslimefun list-threads): Changed return value to also contain + a thread's description. + + * slime.el (slime-list-threads, slime-thread-insert): Adapted to + new return value of LIST-THREADS. + 2008-07-04 Gábor Melis <mega(a)retes.hu> * swank.lisp (call-with-redirected-io): Rebind only standard

1 0

[slime-devel] interrupting REPL-launched thread in background
by alistair gee 05 Jul '08

05 Jul '08

I normally use C-x C-e to execute a form. Since I use either SBCL (with threading enabled) or ACL, slime creates a new thread to execute the form (rather than executing the form in the REPL thread). Occasionally, I need to interrupt or debug the execution. I know that I can use slime-list-threads to see the list of threads. How do I determine which thread is the one that is executing my form? How do I interrupt that thread safely (so that the stack is unwound cleanly)?

3 2

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 05 Jul '08

05 Jul '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1360 slime/ChangeLog:1.1365 --- slime/ChangeLog:1.1360 Wed Jul 2 06:02:14 2008 +++ slime/ChangeLog Fri Jul 4 19:30:44 2008 @@ -1,4 +1,35 @@ -2008-07-02 Martin Simmons <martin(a)lispworks.com> +2008-07-04 Gábor Melis <mega(a)retes.hu> + + * swank.lisp (call-with-redirected-io): Rebind only standard + streams if *GLOBALLY-REDIRECT-IO*. Fixes lost output after + disconnect, reconnect. + +2008-07-04 Willem Broekema <metawilm(a)gmail.com> + + * slime-allegro.lisp (fspec-definition-locations): Workaround for + the issue that Allegro does not record the source file location + for methods defined inside a defgeneric form. The idea is that if + the source location of a method is not found, then the defgeneric + form is almost certainly the right place. + +2008-07-04 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-call-defun): Properly signal error message when + used in a context that is not a function definition. + +2008-07-04 Richard M Kreuter <kreuter(a)progn.net> + + * swank-sbcl.lisp (code-location-source-location), + (code-location-debug-source-name): Patched for incompatible + structure change in SBCL 1.0.18.10. + +2008-07-04 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-call-defun): Broken on DEFMETHOD forms. Fix + that. Also, don't insert package qualifier anymore if the inserted + qualifier is the same as the current REPL package. + +2008-07-02 Martin Simmons <martin(a)lispworks.com> * swank-lispworks.lisp (install-debugger-globally): hook into the environment globally to catch BREAK. @@ -21,8 +52,8 @@ (slime-cl-symbol-package), (slime-qualify-cl-symbol-name): Resurrected from the `slime-parse' contrib, as they've been used by `slime-call-defun'. - -2008-06-02 Raymond Toy <toy.raymond(a)gmail.com> + +2008-06-02 Raymond Toy <toy.raymond(a)gmail.com> Unicode support for CMUCL.

1 0