February 2009 - slime-devel - mailman3.common-lisp.net

Re: [slime-devel] random returning the same value
by Tobias C. Rittweiler 30 Nov '09

30 Nov '09

Patrick May <pjm(a)spe.com> writes: > Hi, > > I'm using Slime in Aquamacs with CCL. When I evaluate (random > 100) repeatedly in CCL started from a shell, I get a randomly > generated number between 0 and 99 inclusive. When I do the same with > C-x C-e within Slime, I get the same number every time. (random 100 > (make- > random-state)) gives the same behavior, but (random 100 (make-random- > state t)) works properly. > > Is Slime caching something or have I hosed up my environment? Sorry for the late response. CCL's swank backend uses threads by default, and threads inherit values from the global environment. *RANDOM-STATE* probably becomes thread-local automatically in CCL, resulting in the behaviour you described. It should work when you type in the form at the Slime REPL which evaluates all forms in the same thread. -T.

4 4

[slime-devel] slime-inspector-copy-down
by Chaitanya Gupta 03 Apr '09

03 Apr '09

Hello, The function slime-inspector-copy-down (bound to M-RET in the inspector) in slime.el has been commented out in cvs for some time. Is it going to be "uncommented" and brought back to life? Or is there an alternative function to do the same thing? Chaitanya

2 1

[slime-devel] unhooking from slime before saving sbcl core?
by Michael Hannemann 02 Mar '09

02 Mar '09

Hi all, I'm new to SLIME, although a co-worker has been prodding me to look at it for a year. I'm also new to SBCL, although I used CMUCL a fair amount several years ago. I apologize if the following is a FAQ -- I scoured the manual and searched through the gmane archives of this list, but I could've missed something. When I write a core image out from SBCL which I'm talking to via SLIME, then try to load it from the command line without SLIME, I get a long stream of warning messages. This happens immediately on startup. I've deleted my .sbclrc file so user init is not to blame. (Also, when I start up SBCL from the shell, this doesn't happen.) The lisp is not dead, just injured; I can get out of the break loop. However, any time I get back into a break loop, I'm in trouble again. My co-worker suggested setting *debugger-hook* to nil, so I tried that, to no avail. I still get the errors (example below), and once I get back to a usable prompt in the lisp, I see that it's set to SWANK:SWANK-DEBUGGER-HOOK again. I've copied the first chunk of what I see on startup to the end of this message. This is MacOSX, SLIME from CVS a day or two ago, SBCL 1.0.23. Any advice? Thanks, Michael --- * Help! 11 nested errors. SB-KERNEL:*MAXIMUM-ERROR-DEPTH* exceeded. 0: (SB-DEBUG::MAP-BACKTRACE #<CLOSURE # {128E6F6D}>)[:EXTERNAL] 1: (BACKTRACE 536870911 #<SYNONYM-STREAM :SYMBOL SB-SYS:*TTY* {128E6DE9}>) [...] 17: (FORMAT #<SB-SYS:FD-STREAM for \"a socket\" {10026331}> \"~6,'0x\")[:EXTERNAL] 18: (SWANK::ENCODE-MESSAGE (:DEBUG 0 11 (\"Couldn't write to #<SB-SYS:FD-STREAM for \\\"a socket\\\" {10026331}>: Bad file descriptor\" \" [Condition of type SB-INT:SIMPLE-STREAM-ERROR]\" NIL) ((\"ABORT\" \"Return to sldb level 10.\") [... and a lot more ...]) 19: (SWANK::SLDB-LOOP 11) 20: (SWANK::SLDB-LOOP 11)[:EXTERNAL] 21: (SWANK::CALL-WITH-BINDINGS ((*PRINT-PRETTY* . T) (*PRINT-LEVEL* . 4) (*PRINT-LENGTH* . 10) (*PRINT-CIRCLE* . T) (*PRINT-READABLY*) (*PRINT-PPRINT-DISPATCH* . #<SB-PRETTY:PPRINT-DISPATCH-TABLE {10224B19}>) (*PRINT-GENSYM* . T) (*PRINT-BASE* . 10) (*PRINT-RADIX*) (*PRINT-ARRAY* . T) (*PRINT-LINES*) (*PRINT-ESCAPE* . T) (*PRINT-RIGHT-MARGIN* . 65) (SWANK::*SLDB-BITVECTOR-LENGTH* . 25) (SWANK::*SLDB-STRING-LENGTH* . 50)) #<FUNCTION (LAMBDA NIL) {10245C55}>) 22: ((LAMBDA (SWANK-BACKEND::DEBUGGER-LOOP-FN)) #<FUNCTION (LAMBDA NIL) {10245C05}>) 23: (SWANK::DEBUG-IN-EMACS #<SB-INT:SIMPLE-STREAM-ERROR {12771809}>) 24: (SWANK:INVOKE-SLIME-DEBUGGER #<SB-INT:SIMPLE-STREAM-ERROR {12771809}>) 25: ((LAMBDA (SWANK-BACKEND::HOOK SWANK-BACKEND::FUN)) #<FUNCTION SWANK:SWANK-DEBUGGER-HOOK> #<CLOSURE (LAMBDA NIL) {12771A2D}>) 26: (SWANK:SWANK-DEBUGGER-HOOK #<SB-INT:SIMPLE-STREAM-ERROR {12771809}> #<unavailable argument>) 27: (INVOKE-DEBUGGER #<SB-INT:SIMPLE-STREAM-ERROR {12771809}>) 28: (INVOKE-DEBUGGER #<SB-INT:SIMPLE-STREAM-ERROR {12771809}>)[:EXTERNAL] [... and on for hundreds of lines ...]

4 4

[slime-devel] abcl special inspector support for java classes, methods, and other good stuff
by Russell McManus 01 Mar '09

01 Mar '09

I have written a couple of dozen lines of code to improve the way the inspector works in abcl for certain java objects. At this point, I can't work without these improvements, but I haven't been able to quite figure out how to produce a proper patch to slime for inclusion. Normally, what i do is execute the following forms after swank is already loaded. This seems to work fine. However when I put all the following forms into swank-abcl.lisp, this approach fails because the swank package is not loaded yet. Any ideas about how I can get organize the following code so that it works cleanly with slime? Thanks, -russ Here is the code: ;; this is the troublesome form. where should it go? (defmethod swank:emacs-inspect ((java-object java-object)) (swank-backend::emacs-inspect-java java-object)) (defun swank-backend::emacs-inspect-java-class (jclass) (flet ((jclass->name (jclass) (let* ((s (jclass-name jclass)) (prefix "java.lang.") (lang-pos (search prefix s))) (if lang-pos (subseq s (+ lang-pos (length prefix))) s)))) (append `("Java Class: " ,(princ-to-string jclass) (:newline)) `("Methods" (:newline)) (loop for method across (jclass-methods jclass) for i = 0 then (1+ i) append (let ((args (mapcar #'jclass->name (coerce (jmethod-params method) 'list)))) `(,(format nil "[~2D] ~A ~A(~{~A~^,~}): ~40T" i (jclass->name (jmethod-return-type method)) (jmethod-name method) args) (:value ,method) (:newline))))))) (defun swank-backend::emacs-inspect-java-object (jobject) (let* ((jclass (jobject-class jobject)) (fields (coerce (jclass-fields jclass) 'list))) (append `("Java Object" ":" ,(princ-to-string jobject) (:newline)) `("Java Class" ":" ,(jclass-name jclass) " " (:value ,jclass) (:newline)) `("Fields" (:newline)) (loop for field in fields for i = 0 then (1+ i) append `(,(format nil "[~2D] ~20A : " i (jfield-name field)) (:value ,(jcall (jmethod (jclass "java.lang.reflect.Field") "get" (jclass "java.lang.Object")) field jobject)) (:newline)))))) (defun swank-backend::emacs-inspect-java-method (jmethod) (let ((return-type (jcall (jmethod (jclass "java.lang.reflect.Method") "getReturnType") jmethod)) (args (coerce (jmethod-params jmethod) 'list))) (setf args (mapcar (lambda (arg) (jclass-name arg)) args)) (append `("Java Method: " ,(jmethod-name jmethod) (:newline)) `("Return Type: " ,(jclass-name return-type) ": " (:value ,return-type) (:newline)) (if args (append `("Arguments" (:newline)) (loop for arg in args for i = 0 then (1+ i) append `(,(format nil "[~2D] ~20A : " i arg) (:value ,arg) (:newline)))) `("Arguments: none" (:newline)))))) (defun swank-backend::emacs-inspect-java (java-object) (flet ((is-a (class-name) (jinstance-of-p java-object (jclass class-name)))) (cond ((is-a "java.lang.Class") (swank-backend::emacs-inspect-java-class java-object)) ((is-a "java.lang.reflect.Method") (swank-backend::emacs-inspect-java-method java-object)) (t (swank-backend::emacs-inspect-java-object java-object)))))

2 3

[slime-devel] Priting values of special variables readable.
by Stas Boukarev 01 Mar '09

01 Mar '09

I find printing values of special variables in echo area with ~A quite misleading, the following patch changes it to ~S. -- With Best Regards, Stas.

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 28 Feb '09

28 Feb '09

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1696 slime/ChangeLog:1.1703 --- slime/ChangeLog:1.1696 Thu Feb 26 23:41:41 2009 +++ slime/ChangeLog Fri Feb 27 22:00:25 2009 @@ -1,5 +1,60 @@ 2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + * slime.el ([test] macroexpand): New test case. + (slime-buffer-visible-p): New helper. + (slime-execute-as-command): New helper. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * hyperspec.el (common-lisp-hyperspec-symbols): Add links to + reader macros. + + * slime.el (slime-reader-macro-at-point): New function. + (slime-hyperspec-lookup): Call it. + + Adapted from Stas Boukarev. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el: Rename `slime-symbol-name-at-point' to + `slime-symbol-at-point'. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el ([portability] lisp-mode-syntax-table): On Emacs21, + make @ a prefix character like it's from Emacs22 + onward. `slime-symbol-name-at-point' was written with that + assumption. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-current-parser-state): Wrap `syntax-ppss' in a + `save-match-data'. This issue has been reported to the Emacs + maintainers. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-defun-if-undefined), + (slime-defmacro-if-undefined): Renamed to + `slime-DEFUN-if-undefined' and `slime-DEFMACRO-if-undefined' to + better differentiate between the two. + + (slime-indulge-pretty-colors): New function. You can now put a + symbol on the plist of `slime-indulge-pretty-colors' to make the + symbol be fontified like `defun'. This is done for + `slime-def-connection-var', and the two symbols above. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-defmacro-if-undefined): New. Analogous to + `slime-defun-if-undefined'. + ([portablity]] with-selected-window) Use it. + ([portability] with-temo-buffer): Likewise. + + Patch by Theam Yong Chew. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + C-c C-c on (defun foo () ,bar) did not result in a compiler note overlay on SBCL. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.178 slime/contrib/ChangeLog:1.184 --- slime/contrib/ChangeLog:1.178 Thu Feb 26 22:44:36 2009 +++ slime/contrib/ChangeLog Fri Feb 27 21:39:32 2009 @@ -1,3 +1,43 @@ +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-arglists.lisp (read-conversatively-for-autodoc): Make it + understand sharpquote forms, so contextual autodoc will work for + `(apply #'foo ...)'. + + * slime-autodoc.el ([test] autodoc.1): New test case, for the + above and more. + (slime-check-autodoc-at-point): New helper. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-parse.el (slime-check-enclosing-form-specs): Use + `slime-test-expect' rather than `slime-check'. + ([test] enclosing-form-specs.1): Add two more cases. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * bridge.el: Replace old-style backquoting. + + * slime-presentations.el + (slime-presentation-sldb-insert-frame-variable-value): Fix typo. + + Patch by Stelian Ionescu. + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-autodoc.el, slime-c-p-c.el, slime-enclosing-context.el + * slime-package-fu.el, slime-parse.el, slime-presentations.el + * slime-xref-browser: Rename `slime-symbol-name-at-point' to + `slime-symbol-at-point' + +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-parse.el (slime-make-extended-operator-parser/look-ahead): + If there's no closing paren (no paredit!) we cannot determine the + end of the list. Check for this. + ([test] enclosing-form-specs.1): Extend test case. + (slime-check-enclosing-form-specs): New helper. + 2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> * swank-fancy-inspector.lisp (emacs-inspect (stream-error)): Do

1 0

[slime-devel] clhs-lookup for reader macros
by Stas Boukarev 27 Feb '09

27 Feb '09

I added reader macros for hyperspec-lookup, hope this would be useful. -- With Best Regards, Stas.

2 1

[slime-devel] [PATCH] a few small fixes
by Stelian Ionescu 27 Feb '09

27 Feb '09

The first patch replaces old-style backquoting with the new-style in contrib/bridge.el(when compiling the file I got a warning that old-style backquotes are deprecated since version 21 and will be removed soon). The other patches fix two typos. -- Stelian Ionescu a.k.a. fe[nl]ix Quidquid latine dictum sit, altum videtur.

2 1

[slime-devel] Patch to fix occasional "jerky" slime-eval-last-expression
by szergling 27 Feb '09

27 Feb '09

Suppose I have 2 windows, 1 for Lisp code, another showing the REPL. Every time I C-xC-e in the Lisp-code buffer, it jerks downwards so that the current line (at point) becomes the last visible line in the window. This is because slime is attempting a (recenter -1), but with the wrong window. I have an older Emacs 21.4.1 here which doesn't have with-selected-window (a macro). The current (the last time I checked) "emulation" of with-selected-window defines a function instead of a macro. This causes evaluation of its body form at the wrong time/places. See eg slime-repl-show-maximum-output (the original source of my investigations and hence the subject of this email). Perhaps a macro equivalent to slime-defun-if-undefined could be used instead? Something like... ;; here, fboundp substitutes for macro-boundp (defmacro slime-defmacro-if-undefined (name &rest rest) `(unless (fboundp ',name) (defmacro ,name ,@rest))) Cheers!

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 27 Feb '09

27 Feb '09

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1687 slime/ChangeLog:1.1696 --- slime/ChangeLog:1.1687 Wed Feb 25 17:53:27 2009 +++ slime/ChangeLog Thu Feb 26 23:41:41 2009 @@ -1,7 +1,72 @@ -2009-02-25 Luís Oliveira <loliveira(a)common-lisp.net> +2009-02-27 Tobias C. Rittweiler <tcr(a)freebits.de> - * contrib/slime-compiler-notes-tree.el: Fix typo in the `provide' - form. + C-c C-c on (defun foo () ,bar) did not result in a compiler note + overlay on SBCL. + + * swank-sbcl.lisp (compiler-note-location): Make it take a + condition; if the condition is a READER-ERROR, the passed + compiler-error-context is very likely NIL---we have not proceeded + beyond reading, so we aren't within the compiler yet. In that + case, we use the stream position of the stream behind the + READER-ERROR instead. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * doc/slime.texi: Fix typos, and add keybindings not listed there. + + Patch by Stas Boukarev. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (sldb-backward-frame): Only move backward when we're + below the backtrace marker. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp (debug-in-emacs): Moved (WITH-BINDINGS + *SLDB-PRINTER-BINDINGS* ...), from here... + (sldb-loop): ... to here. Otherwise results from a user doing an + eval-in-frame were truncated. + + Reported by Jeff Workman. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank.lisp (hash-table-to-alist): New function. + ([method] emacs-inspect (hash-table)): Sort keys if they're all + numbers, symbols, or strings. + + Adapted from Willem Broekema. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-backend.lisp (warn-unimplemented-interfaces): + Bind *PRINT-PRETTY* to T. Otherwise no sugar formatting on CCL. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-pretty-package-name): Signalled an error on + simple symbols; fix that! + + Reported by Geoff Wozniak. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el (slime-current-form-path): Use + `slime-current-parser-state'. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime.el ([test] fancy-symbol-names): New, hopefully + comprehensive, test for funky symbol names. + (slime-check-fancy-symbol-name): Helper. + (slime-exit-vertical-bars): New function to move out from |foo|. + (slime-symbol-constituent-at): New predicate to test whether the + character at point is a valid symbol constituent. + (slime-beginning-of-symbol, slime-end-of-symbol): Rewritten using + above two functions and `forward-sexp' that correctly parses + escapes etc. + (slime-sexp-at-point): Consider thing at point a symbol first. 2009-02-24 Tobias C. Rittweiler <tcr(a)freebits.de> Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.176 slime/contrib/ChangeLog:1.178 --- slime/contrib/ChangeLog:1.176 Sat Feb 21 19:05:21 2009 +++ slime/contrib/ChangeLog Thu Feb 26 22:44:36 2009 @@ -1,3 +1,30 @@ +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-fancy-inspector.lisp (emacs-inspect (stream-error)): Do + not run FILE-POSITION on a closed stream. + +2009-02-26 Tobias C. Rittweiler <tcr(a)freebits.de> + + * slime-parse.el (slime-parse-symbol-name-at-point): + Removed. Superfluous due to recent changes on + `slime-symbol-name-at-point'. + (slime-parse-sexp-at-point): Simplified; use + `slime-sexp-at-point'. + (slime-inside-string-p, slime-beginning-of-string): Use + `slime-current-parser-state'. + ([test] enclosing-form-specs.1): Add some simple cases. + + * slime-enclosing-context.el (slime-find-bound-names): Replace + `slime-parse-symbol-name-at-point' with + `slime-symbol-name-at-point' + (slime-find-bound-functions): No need for `slime-ensure-list' + anymore. + ([test] enclosing-context.1): Adapted due to the changes. + +2009-02-25 Luís Oliveira <loliveira(a)common-lisp.net> + + * slime-compiler-notes-tree.el: Fix typo in the `provide' form. + 2009-02-21 Tobias C. Rittweiler <tcr(a)freebits.de> * slime-package-fu.el: Removed misplaced comma, deleted some

1 0