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

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 14 Sep '04

14 Sep '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.524 slime/ChangeLog:1.528 --- slime/ChangeLog:1.524 Mon Sep 13 23:45:27 2004 +++ slime/ChangeLog Tue Sep 14 23:35:13 2004 @@ -1,3 +1,33 @@ +2004-09-14 Thomas Schilling <tjs_ng(a)yahoo.de> + + * swank-allegro.lisp (inspect-for-emacs): Fixes to previous patch. + +2004-09-14 Marco Baringer <mb(a)bese.it> + + * swank-backend.lisp (inspector, make-default-inspector): Add an + INSPECTOR object argument to the inspector protocol. This allows + implementations to provide more information regarding cretain + objects which can't be, or simply aren't, inspected using the + generic inspector implementation. also export inspect-for-emacs + and related symbols from the backend package. + (make-default-inspector): New function. + + * swank.lisp (inspected-parts): Rename to inspect-for-emacs and + add an inspector argument. Move inspect-for-emacs to + swank-backend.lisp, leave only the default implementations. + + * swank-openml.lisp, swank-sbcl.lisp, swank-allegro.lisp, + swank-cmucl.lisp, swank-lispworks.lisp (inspected-parts): Rename + and change argument list. Many of the inspected-parts methods were + being clobbered by the inspected-parts in swank.lisp, now that + they're being used the return values have been updated for the new + inspect-for-emacs API. + +2004-09-14 Thomas Schilling <tjs_ng(a)yahoo.de> + + * swank-allegro.lisp (inspected-parts): Implement inspector for + structs. + 2004-09-13 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank.lisp (intern-catch-tag): New function.

1 0

[slime-devel] Small typo in the inspector
by Torsten Poulin 14 Sep '04

14 Sep '04

I noticed a small type in the inspector in inspect-for-emacs. My patch also makes it print the word "symbol" without a plural s when applicable. /Torsten Index: swank.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank.lisp,v retrieving revision 1.235 diff -u -r1.235 swank.lisp --- swank.lisp 14 Sep 2004 16:01:51 -0000 1.235 +++ swank.lisp 14 Sep 2004 20:56:06 -0000 @@ -2778,16 +2778,16 @@ (:newline) ,(if (null external-symbols) "0 external symbols." - `(:value ,external-symbols ,(format nil "~D external symbols." (length external-symbols)))) + `(:value ,external-symbols ,(format nil "~D external symbol~:P." (length external-symbols)))) (:newline) ,(if (null internal-symbols) "0 internal symbols." - `(:value ,internal-symbols ,(format nil "~D internals symbols." (length internal-symbols)))) + `(:value ,internal-symbols ,(format nil "~D internal symbol~:P." (length internal-symbols)))) (:newline) ,(if (null (package-shadowing-symbols package)) "0 shadowed symbols." `(:value ,(package-shadowing-symbols package) - ,(format nil "~D shadowed symbols." (length (package-shadowing-symbols package))))))))) + ,(format nil "~D shadowed symbol~:P." (length (package-shadowing-symbols package))))))))) (defmethod inspect-for-emacs ((pathname pathname) (inspector t)) (declare (ignore inspector))

1 0

Re: [slime-devel] MOP problem with clisp
by Thomas Schilling 14 Sep '04

14 Sep '04

Am Tue, 14 Sep 2004 18:54:59 +0200 schrieb marco <mb(a)bese.it>: At least the swank-mop stuff should work this way, but I get some other error (something about the connection). Though I don't have the time to fix it, now.

1 0

Re: [slime-devel] SBCL, multiple (zombie?) processes problems
by rif 14 Sep '04

14 Sep '04

> rif <rif(a)MIT.EDU> writes: > > > Well, on the working machine there was a .swank.lisp explicitly > > setting the communication style to :fd-handler, which I believe is the > > default; both machines were using :fd-handler. > > On SBCL with #+(and SB-THREAD SB-FUTEX) the default is actually :SPAWN. > You're absolutely right. My mistake. Thanks so much for your patience. So I guess :spawn is somewhat problematic on SBCL. I can live with that, since I think I need :fd-handler anyways for my other software that tries to mess with interrupts. Cheers, rif

1 0

[slime-devel] the slime inspector
by marco 14 Sep '04

14 Sep '04

all of this started out with the following goal: it should be _easy_ to ask a live lisp image what's in it, it should be _easy_ to access the associated doc strings and inspect associated "things" like the slots of a class or the symbols in a package. developers should have the option of telling users to inspect the various functions and symbols in a library instead of having to produce api documentation in another form (though that's never a bad idea). the attached patch changes, rather radically, the inspector. the lisp side now deals with formatting and can create "buttons" (certain pieces of text which when clicked call a lambda). this is used to, among other things, implement a simple method deleter. currently it has inspectors for: standard-class, standard-generic-function, standard-method, standard-slot-definiton, array, hash-table, package and cons cells (ciruclar lists, proper lists and "regular" cons cells), features: - access docstrings associated with classes, slots, generic functions, methods, functions and global variables/constants. - makunbound, fmakunbound, remove-method from within inspector. todo: - lots of custom inspectors (strings, ). - improve inspection of arrays (in particular it'd be nice to distinguish between strings, vectors, matrixes and other n-dimensional arrays). - re-implement the list vieweing (don't inspect the entire list at once). - invent some simple syntax for doc strings which will allow the developer to write "See Also" fields and have those "clikable." - distinguish between classes, structs. - types (deftype) - figure out how to distinguish between a defvar/defparameter and defcontstant symbol. - figure out how to access non portable system specific "extras" - compiler-macros. how should this work? - [not sure about this one] have slime-describe-function go to an inspector buffer, have slime-describe-variable inspect the symbol. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen

9 20

[slime-devel] modifying the inspector with implementation specific methods.
by Marco Baringer 14 Sep '04

14 Sep '04

the problem: slime currently loads things in the following order: 1) the backend 2) the lisp system's implementation of the backen 3) swank.lisp initially i've defined everything in swank.lisp, which means that the swank-XYZ.lisp files can add methods but can't redefine the default methods :( (in particular it'd be nice to use the implementation's (inspected-parts T) method). however i can't just define the inspector directly in swank-backend because 1) the inspector uses the swank-mop package which each implementation must define, and 2) that'd be ugly. suggestions? the only thing i can think of is a refactoring of the inspector part of swank.lisp into a general inspector file, one file defining the swank-mop package for each lisp and one file redefining inspector methods for each lisp, but this just feels wrong. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen

3 2

[slime-devel] problems with cross-references
by Giannandrea Castaldi 13 Sep '04

13 Sep '04

Hi, I'm using slime (version of cvs 2004-09-03) with cmucl 19a and I'd like to use the cross-references command but it doesn't work. For example, if I load a file with the following 2 functions: (defun fun1 () (fun2 "pippo")) (defun fun2 (x) (princ x)) And try 'C-c C-W c' on fun2 in the minibuffer there is: 'Who calls: fun2' and after I press return: 'No references found for fun2.' While I expect 'fun1'. Any suggestion? Thanks. Giannandrea

2 1

[slime-devel] catch tags
by Andras Simon 13 Sep '04

13 Sep '04

swank::read-user-input-from-emacs uses numbers as catch tags. Since tags are compared using eq, this looks like a bug. Andras

2 2

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 13 Sep '04

13 Sep '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.517 slime/ChangeLog:1.524 --- slime/ChangeLog:1.517 Sun Sep 12 22:48:30 2004 +++ slime/ChangeLog Mon Sep 13 23:45:27 2004 @@ -1,3 +1,143 @@ +2004-09-13 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> + + * swank.lisp (intern-catch-tag): New function. + (read-user-input-from-emacs, take-input): Use it. + +2004-09-13 John Paul Wallington <jpw(a)gnu.org> + + * swank.lisp (define-special): Make the doc-type `variable' + rather than `symbol'. Don't quote `doc'. Doc fix. + +2004-09-09 Martin Simmons <martin(a)xanalys.com> + + * swank-lispworks.lisp: Set up the swank-mop package. Implement + swank-mop:slot-definition-documentation and function-name. + +2004-09-13 Marco Baringer <mb(a)bese.it> + + * swank.lisp (inspected-parts): Added inspectors for pathnames, + logical pathnames, standard-objects and numbers (float, ratio, + integer and complex). + + * swank-backend.lisp: Define import-to-swank-mop. + + * swank-openmcl.lisp, swank-sbcl.lisp, swank-allegro.lisp: Don't + define the import-to-swank-mop function (now defined in + swank-backend.lisp). + + * swank-cmucl.lisp (swank-mop, function-name): Implement backend + for inspector. + (arglist): Add support for extracting arglists from function + objects. + (create-socket): Don't specify the host on PPC. + +2004-09-13 Alan Ruttenberg <alanr-l(a)mumble.net> + + * slime.el slime-goto-location-position: New location specifiers: + (:method name specializers . qualifiers) all are strings. Looks + for defxxx name then the qualifiers as words, in order then the + specializers as words, in order (except for "T", which is + optional). Pass the symbols names for specializers and qualifiers + (no packages). Used by openmcl but might be useful for others + (:text-anchored <position fixnum> <string> <delta fixnum>) + Got to position, then search for string, then move delta. To + support upcoming source recording for openmcl debugging. + + * swank-openmcl multiple changes: - fix support for *sldb-top* + (formerly *swank-debugger-stack-frame*) Was not thread safe. Now + (application-error), and (interrupt-thread) records the error + pointer in a table associated with thread and map-backtrace picks + up the appropriate pointer. *process-to-stack-top*, + (grab-stack-top), (record-stack-top). + + - Other adjustments for changes to multiprocessing: remove + (force-break-in-listener) no longer necessary since we use + process-interrupt instead of ccl::*interactive-abort-process* + Adjust break-in-sldb to do so for swank repl connections + (abstraction breaking reference to swank::*connections*, but + nicely via intern) + + - changes to (find-definitions) (function-source-location), + addition of (maybe-method-location) (remove-filename-quoting). To support + editing definitions of methods. To fix bug with pathnames with + quoted characters (like "\\.swank.lisp"). To remove bogus source + recording of l1-boot-3 in functions that didn't have a source file + noted. + + - Implementation of xref functions: (xref-locations) uses xref + implementation added to openmcl recently. Note that you have to + (ccl::start-xref) for it to work for other than who-calls, and + that xref information is not currently persisted in fasl files (I + will release a patch for this soon) Backend functions (who-binds) + (who-macroexpands) (who-references) (who-sets) + (who-calls) (list-callees) (who-specializes) + + - Lifted profile backend functions from swank-clisp which use + "metering.lisp" + + - (openmcl-set-debug-switches) turns on the various variables I. + know about that have the lisp record extra debugging + information(including starting xref). I suggest you call + it. Should it be called by default? + + - (frame-arguments) use builtin ccl::frame-supplied-args since the + current version was sometimes missing the first argument to the + function. (I think this was when it was passed by register. If you + don't want to lose it in the frame locals in backtrace, call + (openmcl-set-debug-switches) specifically, set + ccl::*ppc2-compiler-register-save-label* to t + + - implement frame-var-value backend + + * metering.lisp: Minor changes to #+ #- to recognize openmcl + + * swank-loader.lisp: Load "metering.lisp" + +2004-09-13 Marco Baringer <mb(a)bese.it> + + * swank.lisp (inspected-parts): Deal with unfinalized classes in + standard-class inspector. (Patch from Thomas Schilling) + +2004-09-13 Marco Baringer <mb(a)bese.it> + + * swank.lisp: New inspector protocol. The lisp side now returns a + specially formated list of "things" to format which are then + passed to emacs and rendered in the inspector buffer. Things can + be either text, recursivly inspectable values, or functions to + call. + (inspected-parts): Redefine for new inspector protocol. + (*inspectee-parts*): Redefine as array. + (*inspectee-actions*): New array, similar to *inspectee-parts*. + (reset-inspector): Update for new implementation of + *inspectee-parts* and new variable *inspectee-actions*. + (inspector-contents-for-emacs): New function. + (inspect-object): Update for new inspector protocol. + (inspector-nth-part): Update for new *inspectee-parts* + implementation. + (inspector-call-nth-action): New function. + + * slime.el (slime-inspector-action-face): New face. + (slime-open-inspector): Adapt to new inspector protocol. + (slime-inspector-operate-on-point): New function, subsumes + slime-inspector-inspect-object-at-point. + (slime-inspector-next-inspectable-object): Skip to next object, + not just end of current object; wrap around buffer. + (slime-inspector-mode-map): change bindings of [return] and "\C-m" + + * swank-bacend.lisp (swank-mop): New package. Simply defines all + the MOP related symbols we need from an implementation. + (arglist): Update doc string. Provide default implementation. + (function-name): New backend function. + + * swank-allegro.lisp (swank-mop, slot-definition-documentation): + Implement. (Patch from Thomas Schilling) + + * swank-sbcl.lisp (swank-mop, slot-definition-documentation, + function-name): Implement. + + * swank-openmcl.lisp (swank-mop, function-name): Implement. + (arglist): Implement for function objects. + 2004-09-12 Helmut Eller <e9626484(a)stud3.tuwien.ac.at> * swank.lisp (compile-file-for-emacs): Use with-buffer-syntax so

1 0

[slime-devel] define-special fix
by John Paul Wallington 13 Sep '04

13 Sep '04

It seems that the doc-type in the (setf documentation) call should be `variable' rather than `symbol'. Also, it doesn't seem necessary to quote `doc'. WDYT ? 2004-09-06 John Paul Wallington <jpw(a)gnu.org> * swank.lisp (define-special): Make the doc-type `variable' rather than `symbol'. Don't quote `doc'. Index: swank.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank.lisp,v retrieving revision 1.227 diff -u -r1.227 swank.lisp --- swank.lisp 3 Sep 2004 21:12:00 -0000 1.227 +++ swank.lisp 6 Sep 2004 03:15:30 -0000 @@ -905,7 +905,7 @@ This is like defvar, but NAME will not initialized." `(progn (defvar ,name) - (setf (documentation ',name 'symbol) ',doc))) + (setf (documentation ',name 'variable) ,doc))) (define-special *buffer-package* "Package corresponding to slime-buffer-package.

2 2