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

[slime-devel] Swank broken in ECL
by bojohan+news＠dd.chalmers.se 06 Dec '06

06 Dec '06

;;; Loading "/home/bojohan/cvs/slime/swank.lisp" Iteration in LOOP follows body code. Current LOOP context: FOR EL = (AREF VECTOR IDX) DO. Broken at LOAD. SWANK>> 2006-12-06 Johan Bockgård <bojohan+sf(a)dd.chalmers.se> * swank.lisp (fuzzy-completion-set): Don't mix for clauses and body clauses in loop. --- swank.lisp 06 Dec 2006 10:06:07 +0100 1.420 +++ swank.lisp 06 Dec 2006 11:46:39 +0100 @@ -3460,8 +3460,7 @@ (parse-completion-arguments string default-package-name) (flet ((convert (vector &optional converter) (when vector - (loop for idx :upfrom 0 - while (< idx (length vector)) + (loop for idx below (length vector) for el = (aref vector idx) do (setf (aref vector idx) (convert-fuzzy-completion-result el converter internal-p package-name))))))

2 1

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

06 Dec '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1010 slime/ChangeLog:1.1013 --- slime/ChangeLog:1.1010 Mon Dec 4 23:46:06 2006 +++ slime/ChangeLog Tue Dec 5 09:36:34 2006 @@ -1,4 +1,22 @@ +2006-12-05 Helmut Eller <heller(a)common-lisp.net> + + * swank.lisp (create-swank-server): Removed. Use create-server + instead. + + * slime.el (slime-first-change-hook): Don't do anything if buffers + file doesn't exist. + (slime-start, slime-set-connection-info): Add support for a + :init-function which is called after the usual initialization of the + connection is completed. + + * swank-source-file-cache.lisp (buffer-first-change): Always + return nil and remove the now redundant test with probe-file. + + * swank-backend.lisp (guess-external-format): Return nil if the + file can't be opened. Previusly we wrongly read from stdin. + 2006-12-05 Juho Snellman <jsnell(a)iki.fi> + Real xref support for SBCL (requires SBCL 1.0.0.18). * swank-sbcl.lisp (who-calls): New function, fetch xref data from

1 0

[slime-devel] SLIME can not connect to running SBCL
by vedm 05 Dec '06

05 Dec '06

Latest version of SBCL on Debian, latest SLIME on CVS: I have the image running, try slime-connect, and nothing happens. It doesn't matter if SBCL is running under attachtty or not. When I start SBCL with SLIME (with the "slime" Emacs function), then it starts, but the start-up process stops several times and continues only after I press some key - any key. Any clue? -- vedm

3 5

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 05 Dec '06

05 Dec '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1009 slime/ChangeLog:1.1010 --- slime/ChangeLog:1.1009 Sun Nov 26 13:08:30 2006 +++ slime/ChangeLog Mon Dec 4 23:46:06 2006 @@ -1,3 +1,28 @@ +2006-12-05 Juho Snellman <jsnell(a)iki.fi> + Real xref support for SBCL (requires SBCL 1.0.0.18). + + * swank-sbcl.lisp (who-calls): New function, fetch xref data from + sb-introspect. + (who-binds): Ditto. + (who-sets): Ditto. + (who-references): Ditto. + (who-macroexpands): Ditto. + (defxref): New macro, create the above functions. + (source-location-for-xref-data): New, map from sb-introspect xref + format to the Swank xref format. + (sanitize-xrefs): Map PCL method names to something more readable. + (string-path-snippet): New function, finds a more accurate source + snippet for definition source locations which have both an + :emacs-string and a full source path available. Otherwise the xref + location would point to the toplevel form rather than the exact + form for functions compiled with C-c C-c. + (source-file-position): New function, somewhat like + source-path-file-position but uses the source-file cache, handles + missing form-paths more gracefully. + (make-definition-source-location): Use the above two functions. + (sbcl-with-xref-p): New function, detect whether SBCL has xref support + for backwards compability. + 2006-11-26 Juho Snellman <jsnell(a)iki.fi> * swank-source-file-cache.lisp (buffer-first-change): Check

1 0

[slime-devel] slime-repl-matching-input-regexp
by Ivan Shvedunov 04 Dec '06

04 Dec '06

Hi. I've sent a patch some time ago that made it possible to use paredit in REPL window without breaking history. It changed the behavior of slime-repl-matching-input-regexp so that the text between the start of current REPL input and the caret is used for history search. The patch was applied on 2006-09-01 by Nikodemus Siivola but later reverted due to some problems that it caused (there was a vague reference to complaints on #lisp). Well, now I had to repatch my copy of slime.el... No big problem with that, but I just wonder what were these negative consequences ? Maybe slime-repl-matching-input-regexp behavior should be made customizable?.. Ivan

2 1

[slime-devel] documentation
by Jeffrey Cunningham 30 Nov '06

30 Nov '06

I just committed changes to doc/slime.texi in CVS. These changes complete the addition of all (current) external functions to the functions index. They are linked to their descriptions. Where there are currently no descriptions I have put them in either slime-mode/Commands/Miscellaneous commands or SLDB: the .../Miscellaneous commands until such time as I or someone else can add descriptions and move them into more suitable locations. If I've left any out, or included (probably) some that shouldn't be there, please let me know. I'm still thinking about how one might do key sub-indices. Regards, --Jeff

1 0

[slime-devel] a per-dialect -connected-hook?
by _deepfire＠feelingofgreen.ru 28 Nov '06

28 Nov '06

Would a patch implementing per-slime-dialect slime-dialect-*-connected-hook by hooking a dispatcher onto slime-connected-hook be deemed a good idea? regards, Samium Gromoff

1 0

[slime-devel] Re: [slime-cvs] CVS slime
by Helmut Eller 28 Nov '06

28 Nov '06

* jsnellman [2006-11-26 19:08+0100] writes: > Log Message: > * swank-source-file-cache.lisp (buffer-first-change): Check > whether a file exists before trying load it into the source > cache. [...] > (defimplementation buffer-first-change (filename) > "Load a file into the cache when the user modifies its buffer. > This is a win if the user then saves the file and tries to M-. into it." > - (unless (source-cached-p filename) > - (ignore-errors (source-cache-get filename (file-write-date filename))))) > + (unless (or (source-cached-p filename) > + (not (ignore-errors (probe-file filename)))) > + (ignore-errors > + (source-cache-get filename (file-write-date filename))))) > I don't understand this change. If the file doesn't exist then OPEN will just fail (which is already handled). And even if PROBE-FILE succeeds, OPEN can still fail, e.g. because some other process deleted the file in between. Which problem does the patch solve?

2 1

[slime-devel] `slime-repl-abort-read' bug?
by Taylor R Campbell 27 Nov '06

27 Nov '06

`slim-repl-abort-read' ignores its two arguments and just pops `slime-read-string-threads' and `slime-read-string-tags'. Is the Lisp system really assumed to request and to abort REPL reads in a LIFO manner, or is this a bug? Or is my understanding of the way SLIME's REPL read mode works bogus?

2 3

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 27 Nov '06

27 Nov '06

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1007 slime/ChangeLog:1.1009 --- slime/ChangeLog:1.1007 Fri Nov 24 18:34:58 2006 +++ slime/ChangeLog Sun Nov 26 13:08:30 2006 @@ -1,3 +1,21 @@ +2006-11-26 Juho Snellman <jsnell(a)iki.fi> + + * swank-source-file-cache.lisp (buffer-first-change): Check + whether a file exists before trying load it into the source cache. + +2006-11-26 Juho Snellman <jsnell(a)iki.fi> + + Restore the way M-n and M-p used to work in the REPL. (cherry-picked + from a patch with other changes, sent by Attila Lendvai). + + * slime.el (slime-repl-previous-input-starting-with-current-input) + (slime-repl-next-input-starting-with-current-input): New functions, + work like the old slime-repl-previous-input / next-input. + (slime-repl-matching-input-regexp): Restore old version. + (slime-repl-mode-map): Bind s-r-p-i-s-w-c-i and s-r-n-i-s-w-c-i + to M-p and M-n respectively. slime-repl-previous-input and + slime-repl-next-input are still accessible with C-up / C-down. + 2006-11-25 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime-repl-read-break): Use a :emacs-interrupt message

1 0