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

[slime-devel] Daily ChangeLog diff
by Helmut Eller 14 Jun '11

14 Jun '11

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.480 slime/contrib/ChangeLog:1.481 --- slime/contrib/ChangeLog:1.480 Sat Jun 11 12:24:36 2011 +++ slime/contrib/ChangeLog Tue Jun 14 02:33:39 2011 @@ -1,3 +1,11 @@ +2011-06-14 Nikodemus Siivola <nikodemus(a)random-state.net> + + * slime-cl-indent.el (lisp-beginning-of-defmethod): New function. + (lisp-indent-defmethod): Extend to work with non-toplevel defmethods, + and method definitions inside defgenerics. + + * slime-cl-indent-test.txt: Tests 40-43. + 2011-06-11 Nikodemus Siivola <nikodemus(a)random-state.net> * slime-cl-indent.el (common-lisp-style-names): XEmacs doesn't

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 12 Jun '11

12 Jun '11

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.477 slime/contrib/ChangeLog:1.480 --- slime/contrib/ChangeLog:1.477 Fri Jun 10 13:03:33 2011 +++ slime/contrib/ChangeLog Sat Jun 11 12:24:36 2011 @@ -1,3 +1,36 @@ +2011-06-11 Nikodemus Siivola <nikodemus(a)random-state.net> + + * slime-cl-indent.el (common-lisp-style-names): XEmacs doesn't + like lists either in completing-read -- need an alist. + + XEmacs compatibility. + + * slime-cl-indent.el (common-lisp-style-names): New function. + (common-lisp-set-style): Call common-lisp-style-names instead of + using the hashtable directly: XEmacs doesn't like hash-tables in + completing-read. + (common-lisp-looking-back): New function. + (common-lisp-indent-function-1): Use common-lisp-looking-back instead + of looking-back: XEmacs doesn't have looking-back. + + Refactoring named styles. + + * slime-cl-indent.el (define-common-lisp-style): Not a list of hooks anymore, + but just a single hook. Document the fact that code in :eval option may get + called multiple times over the lifetime of a buffer. + (common-lisp-style-name, common-lisp-style-inherits) + (common-lisp-style-variables, common-lisp-style-indentation) + (common-lisp-style-hook, common-lisp-style-docstring) + (common-lisp-make-style, common-lisp-find-style): New functions. + Convenience accessors, locator, and constructor. + (common-lisp-add-style): Don't precompute inheritance. + (common-lisp-safe-style-p): For use with safe-local-variable. + (common-lisp-activate-style): Now handles inheritance. + (common-lisp-active-style-methods): New function, computes and caches + inheritance. + (common-lisp-set-style): Changes to match others. + (common-lisp-get-indentation): Use common-lisp-active-style-methods. + 2011-06-10 Nikodemus Siivola <nikodemus(a)random-state.net> Boa-constructor indentation.

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 11 Jun '11

11 Jun '11

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2197 slime/ChangeLog:1.2198 --- slime/ChangeLog:1.2197 Thu Jun 9 10:48:47 2011 +++ slime/ChangeLog Fri Jun 10 09:13:28 2011 @@ -1,3 +1,9 @@ +2011-06-10 Nikodemus Siivola <nikodemus(a)random-state.net> + + * swank.lisp (macro-indentation): Walk the lambda-list + to construct a better indentation spec instead of just + looking for &BODY. + 2011-06-09 Nikodemus Siivola <nikodemus(a)random-state.net> * swank-sbcl.lisp (execv): Stupid SBCL hackers breaking backwards Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.470 slime/contrib/ChangeLog:1.477 --- slime/contrib/ChangeLog:1.470 Thu Jun 9 13:47:46 2011 +++ slime/contrib/ChangeLog Fri Jun 10 13:03:33 2011 @@ -1,3 +1,51 @@ +2011-06-10 Nikodemus Siivola <nikodemus(a)random-state.net> + + Boa-constructor indentation. + + * slime-cl-indent.el: Give :constructor the appropriate + indentation method. + + * slime-cl-indent-test.txt: Test-case 39. + + * slime-cl-indent.el (common-lisp-indent-function-1): Missing + paren. Teaches me to to paredit. + + Take , and ,@ into account properly. + + * slime-cl-indent.el (common-lisp-indent-function-1): Adjust + normal-indent to take , and ,@ into account + + * slime-cl-indent-test.txt: Test case 36-38. + + Better handling of complex indentation specs. + + * slime-cl-indent.el (common-lisp-run-indentation-tests): Add + a style with complex specs for testing. + (lisp-indent-259): fix handling of nested &whole. Hopefully. + + slime-indentation: Comment indentation in styles. + + * slime-cl-indent.el ("basic"): Hack comment indentation + to avoid inserting spaces where they don't belong. + + slime-indentation: More adjustment to the fallback method. + + * slime-cl-indent.el (common-lisp-indent-function-1): handle + case of empty or comment line following the opening paren + of the containing expression. + + * slime-cl-indent-test.txt: new test case. + + slime-indentation: Test refactoring. + + * slime-cl-indent.el (common-lisp-indent-test): replaces + test-lisp-indent. + (common-lisp-run-indentation-tests): replaces + run-lisp-indent-tests. + + * slime-cl-indent-test.txt: new file, contains all the indentation + tests. + 2011-06-09 Nikodemus Siivola <nikodemus(a)random-state.net> * slime-cl-indent.el (lisp-lambda-list-keyword-parameter-alignment):

1 0

[slime-devel] slime-indentation note
by Nikodemus Siivola 10 Jun '11

10 Jun '11

I've now committed all my pending patches to slime-indentation, and I think I'm mostly done with it for the duration -- modulo bugs. If anyone runs across bogus indentation behaviour, please do let me know. If it's a regression, I'll try to deal with it ASAP. If it's something that's always been bogus... depend on how hard it looks. I would like to propose adding slime-indentation to slime-fancy after giving it a few more weeks to gather dust in case of surprises. Any objections to that? Cheers, -- nikodemus

2 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 10 Jun '11

10 Jun '11

Index: slime/ChangeLog diff -u slime/ChangeLog:1.2195 slime/ChangeLog:1.2197 --- slime/ChangeLog:1.2195 Fri May 27 00:45:44 2011 +++ slime/ChangeLog Thu Jun 9 10:48:47 2011 @@ -1,3 +1,24 @@ +2011-06-09 Nikodemus Siivola <nikodemus(a)random-state.net> + + * swank-sbcl.lisp (execv): Stupid SBCL hackers breaking backwards + compatibility again: SB-POSIX:SYSCALL-ERROR has a required + argument in 1.0.49, but accepts no arguments in older ones. Use + plain ERROR instead. + + Support for per-package derived indentation, when + slime-indentation is used. + + * swank.lisp (update-indentation/delta-for-emacs): Tell Emacs + which packages the symbol is available in. Unless + slime-indentation is used, this information is just dropped on the + floor. + + * slime.el (slime-update-system-indentation): New function. Use + this to inform indentation about derived specs when + `common-lisp-system-indentation' is bound. + (slime-handle-indentation-update): Adjust to support per-package + derived indentation specs when slime-indentation is available. + 2011-05-27 Helmut Eller <heller(a)common-lisp.net> Fix "wrong number of args" problem with slime-inspector-quit. Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.466 slime/contrib/ChangeLog:1.470 --- slime/contrib/ChangeLog:1.466 Wed Jun 8 14:04:53 2011 +++ slime/contrib/ChangeLog Thu Jun 9 13:47:46 2011 @@ -1,5 +1,38 @@ 2011-06-09 Nikodemus Siivola <nikodemus(a)random-state.net> + * slime-cl-indent.el (lisp-lambda-list-keyword-parameter-alignment): + fix docstring: key1 and key3 were aligned in source, but not in *Help*. + (common-lisp-delete-style): Missing function, needed by tests. + (common-lisp-indent-function-1, run-lisp-indent-tests): Fix fallback + to handle case of first function argument on its own line. + + Support for per-package derived indentation. + + * slime-cl-indent.el (common-lisp-system-indentation) + (common-lisp-guess-current-package) + (common-lisp-current-package-function) + (common-lisp-symbol-package): New variables and functions. + (common-lisp-get-indentation): Adjust to use system derived + information when available. + (common-lisp-indent-function-1): Adust to provide + `common-lisp-get-indentation' with the full symbol incl. + package prefix. + + * slime-indentation.el (common-lisp-current-package-function): set + to `slime-current-package'. + + Support for named styles. + + * slime-cl-indent.el (common-lisp-style) + (common-lisp-style-default, common-lisp-active-style) + (common-lisp-lisp-mode-hook, common-lisp-styles) + (common-lisp-add-style, define-common-lisp-style) + (common-lisp-set-style, common-lisp-set-style-history) + (common-lisp-activate-style, common-lisp-get-indentation): New + functions and variables. Setting `common-lisp-style' buffer local + variable in a file picks a style -- or use `common-lisp-set-style" + to do it. Indentation from style takes precedence. + Indirect indentation specs. * slime-cl-indent.el (common-lisp-indent-function)

1 0

[slime-devel] SWANK patch for SBCL 1.0.49
by Stelian Ionescu 09 Jun '11

09 Jun '11

1.0.48.27 added a required argument to sb-posix:syscall-error -- Stelian Ionescu a.k.a. fe[nl]ix Quidquid latine dictum sit, altum videtur. http://common-lisp.net/project/iolib

2 1

[slime-devel] slime-indentation: common-lisp-style
by Nikodemus Siivola 09 Jun '11

09 Jun '11

I've added a tentative implementation of indentation styles to slime-indentation on my Github tree: https://github.com/nikodemus/Slime/tree/slime-indentation With this, you can do ;; -*- common-lisp-style: stylename -*- which will then set the indentation specs and buffer-local variables specified for that style in common-lisp-style-alist. (Or you can do M-x set-common-lisp-style as well, of course.) The big deal is that with this you can have eg. IF ident with 2 spaces in one file, and with 4 in another -- as fits the project style. Without this I need to have multiple emacses started with different command-line options which is a huge pain -- and I expect I'm not the only one. Here's a sample style: (push '("eccentric" (variables (lisp-lambda-list-keyword-alignment t) (lisp-lambda-list-keyword-parameter-alignment t) (lisp-loop-indent-subclauses nil) (lisp-loop-indent-forms-like-keywords t)) (indentation (if (4 2 2)))) common-lisp-style-alist) Things I'd like to sort out before proposing this for Slime CVS: - tests - support for inheriting styles - define-common-lisp-style so we're not married to the alist spec - get someone with more emacs-fu to review this ...the last in particular I would appreciate help with. I've no idea if the things I'm doing with buffer-local variables are kosher or not. In particular, I feel a bit icky about (add-hook 'lisp-mode-hook (lambda () (hack-local-variables))) which I needed to do because (lisp-mode) kills all local variables. Cheers, -- nikodemus

1 1

[slime-devel] Daily ChangeLog diff
by Helmut Eller 09 Jun '11

09 Jun '11

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.464 slime/contrib/ChangeLog:1.466 --- slime/contrib/ChangeLog:1.464 Sat Jun 4 11:20:15 2011 +++ slime/contrib/ChangeLog Wed Jun 8 14:04:53 2011 @@ -1,3 +1,16 @@ +2011-06-09 Nikodemus Siivola <nikodemus(a)random-state.net> + + Indirect indentation specs. + + * slime-cl-indent.el (common-lisp-indent-function) + (common-lisp-get-indentation, common-lisp-indent-function-1) + (lisp-indent-defmethod): Indirect using (as foo) -style specs. + +2011-06-08 Nikodemus Siivola <nikodemus(a)random-state.net> + + * slime-cl-indent.el (common-lisp-indent-function-1): fallback + method to deal with trailing expressions on the previous line. + 2011-06-04 Nikodemus Siivola <nikodemus(a)random-state.net> * slime-cl-indent.el (toplevel):

1 0

[slime-devel] Daily ChangeLog diff
by Helmut Eller 05 Jun '11

05 Jun '11

Index: slime/contrib/ChangeLog diff -u slime/contrib/ChangeLog:1.463 slime/contrib/ChangeLog:1.464 --- slime/contrib/ChangeLog:1.463 Tue May 31 03:30:02 2011 +++ slime/contrib/ChangeLog Sat Jun 4 11:20:15 2011 @@ -1,3 +1,9 @@ +2011-06-04 Nikodemus Siivola <nikodemus(a)random-state.net> + + * slime-cl-indent.el (toplevel): + (run-lisp-indent-tests): Better named-lambda and + destructuring-bind indenation. + 2011-05-31 Stas Boukarev <stassats(a)gmail.com> * slime-package-fu.el (slime-export-symbol-representation-auto):

1 0

[slime-devel] help with Windows XP remote connection to swank
by George Oliver 02 Jun '11

02 Jun '11

hi, apologies that this is crossposted to emacs.help as I afterward found this list. I'm trying to connect to a remote swank from my local native Windows Emacs Slime. This seems to be possible but I haven't yet found a working method. Here's what I'm doing so far: 1. open a ssh tunnel to remote server C:\Documents and Settings\george\My Documents>plink -P <ssh port> -i "<private key file>" -ssh -L 4005:localhost:<ssh port> <user>@<host> 2. start swank server on remote server by logging in via ssh Authenticating with public key "imported-openssh-key" Passphrase for key "imported-openssh-key": Last login: Thu Jun 2 00:38:06 2011 from <address> gmk@labs:~$ ccl64 Welcome to Clozure Common Lisp Version 1.6-r14468M (LinuxX8664)! ? (ql:quickload "swank") To load "swank": Load 1 ASDF system: swank ; Loading "swank" ("swank") ? (swank:create-server) ;; Swank started at port: 4005. 4005 3. M-x slime-connect in local Emacs and accept defaults This returns the following error: Debugger entered--Lisp error: (cl-assertion-failed (plusp length)) (condition-case error (slime-net-read) (error (debug ... error) (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> "Protocol mismatch.\n") recursive-edit() byte-code("\306 @\307=\203! (note that it seems impossible to cut and paste that text, I guess from whatever control characters it's outputting to the buffer -- however the cl-assertion-failed seems to be the main problem, for reference see http://thread.gmane.org/gmane.lisp.slime.devel/4732 <http://www.google.com/url?sa=D&q=http://thread.gmane.org/gmane.lisp.slime.d…>). While I had hopes for this guide (http:// webcache.googleusercontent.com/search? q=cache:C3LnVLiH5lkJ:www.a1k0n.net/blah/archives/2005/11/index.html+ <http://www.google.com/url?sa=D&q=www.a1k0n.net/blah/archives/2005/11/index.…> %22Using+SLIME+over+an+SSH+tunnel %22&cd=1&hl=en&ct=clnk&gl=us&source=www.google.com <http://www.google.com/url?sa=D&q=www.google.com>), it doesn't seem like any of the plink-based methods use key authentication. I'm also having problems connecting just with tramp, as I can't get plink key authentication to work with that method. There doesn't seem to be a good way to specify the key. Specifying a saved session with plink doesn't work, i.e. it returns Unable to open connection: Host does not exist Though normal use of plink works fine. thanks for any help, George

1 1