slime-devel

slime-devel@common-lisp.net

4231 discussions

[slime-devel] some inspector conveniencies (patch)
by Thomas Schilling 25 Oct '04

25 Oct '04

I think I already sent these some days ago but they never got commited. So I revised them and resend them, in short it's (all concerning the inspector): - when inspecting gf: - display short docs for gf methods (when available) - sort them by applicability (according to a #lisp suggestion some time ago), although this is still a little buggy but IMO still better than no ordering at all users can specify their own function if they don't like the default by setting `*gf-method-getter*' (feel free to rename this to sth. more sensible) - when inspecting class: - abbrev'd docs for methods specialising on the class - print them one per line - order them by name (not customizable) -ts ,---- Index: swank.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank.lisp,v retrieving revision 1.253 diff -u -r1.253 swank.lisp --- swank.lisp 19 Oct 2004 06:14:17 -0000 1.253 +++ swank.lisp 24 Oct 2004 19:48:01 -0000 @@ -2786,6 +2786,68 @@ collect "#<unbound>" collect '(:newline))))) +(defvar *gf-method-getter* 'methods-by-applicability + "This function is called to get the methods of a generic function. +The default returns the method sorted by applicability. +See `methods-by-applicability'.") + +;;; Largely inspired by (+ copied from) the McCLIM listener +(defun methods-by-applicability (gf) + "Return methods ordered by qualifiers, then by most specific argument types. + +Qualifier ordering is: :before, :around, primary, and :after. +We use the length of the class precedence list to determine which type is +more specific." + ;;FIXME: How to deal with argument-precedence-order? + (let* ((methods (copy-list (swank-mop:generic-function-methods gf))) + (lambda-list (swank-mop:generic-function-lambda-list gf))) + ;; sorter function (most specific is defined as smaller) + (flet ((method< (meth1 meth2) + ;; First ordering rule is by qualifiers, that is :before-methods + ;; come before :around methods, before primary methods, before + ;; :after methods, other qualifiers are treated like none at all + ;; (so like primary methods) + (let ((qualifier-order '(:before :around nil :after))) + (let ((q1 (or (position (first (swank-mop:method-qualifiers meth1)) qualifier-order) 2)) + (q2 (or (position (first (swank-mop:method-qualifiers meth2)) qualifier-order) 2))) + (cond ((< q1 q2) (return-from method< t)) + ((> q1 q2) (return-from method< nil))))) + ;; If qualifiers are equal, go by arguments + (loop for sp1 in (swank-mop:method-specializers meth1) + for sp2 in (swank-mop:method-specializers meth2) + do (cond + ((eq sp1 sp2)) ;; continue comparision + ;; an eql specializer is most specific + ((typep sp1 'swank-mop:eql-specializer) + (return-from method< t)) + ((typep sp2 'swank-mop:eql-specializer) + (return-from method< nil)) + ;; otherwise the longer the CPL the more specific + ;; the specializer is + ;; FIXME: Taking the CPL as indicator has the problem + ;; that unfinalized classes are most specific. Can we pick + ;; a reasonable default or do something with SUBTYPEP ? + (t (let ((l1 (if (swank-mop:class-finalized-p sp1) + (length (swank-mop:class-precedence-list sp1)) + 0)) + (l2 (if (swank-mop:class-finalized-p sp2) + (length (swank-mop:class-precedence-list sp2)) + 0))) + (cond + ((> l1 l2) + (return-from method< t)) + ((< l1 l2) + (return-from method< nil)))))) + finally (return nil)))) + (declare (dynamic-extent #'method<)) + (sort methods #'method<)))) + +(defun abbrev-doc (doc &optional (maxlen 80)) + "Return the first sentence of DOC, but not more than MAXLAN characters." + (subseq doc 0 (min (1+ (or (position #\. doc) (1- maxlen))) + maxlen + (length doc)))) + (defmethod inspect-for-emacs ((gf standard-generic-function) (inspector t)) (declare (ignore inspector)) (values "A generic function." @@ -2797,13 +2859,17 @@ "It uses " (:value ,(swank-mop:generic-function-method-combination gf)) " method combination." (:newline) "Methods: " (:newline) ,@(loop - for method in (swank-mop:generic-function-methods gf) + for method in (funcall *gf-method-getter* gf) collect `(:value ,method ,(inspector-princ ;; drop the first element (the name of the generic function) (cdr (method-for-inspect-value method)))) collect " " collect (let ((meth method)) `(:action "[remove method]" ,(lambda () (remove-method gf meth)))) + collect '(:newline) + if (documentation method t) + collect " Documentation: " and + collect (abbrev-doc (documentation method t)) collect '(:newline))))) (defmethod inspect-for-emacs ((method standard-method) (inspector t)) @@ -2860,8 +2926,18 @@ ,@(when (swank-mop:specializer-direct-methods class) `("It is used as a direct specializer in the following methods:" (:newline) ,@(loop - for method in (swank-mop:specializer-direct-methods class) + for method in (sort (copy-list (swank-mop:specializer-direct-methods class)) + #'string< :key (lambda (x) + (symbol-name + (let ((name (swank-mop::generic-function-name + (swank-mop::method-generic-function x)))) + (if (symbolp name) name (second name)))))) + collect " " collect `(:value ,method ,(inspector-princ (method-for-inspect-value method))) + collect '(:newline) + if (documentation method t) + collect " Documentation: " and + collect (abbrev-doc (documentation method t)) and collect '(:newline)))) "Prototype: " ,(if (swank-mop:class-finalized-p class) `(:value ,(swank-mop:class-prototype class))

2 1

[slime-devel] tiny inspector fix for acl62
by Thomas Schilling 24 Oct '04

24 Oct '04

I don't like error messages ... ,------ Index: swank-allegro.lisp =================================================================== RCS file: /project/slime/cvsroot/slime/swank-allegro.lisp,v retrieving revision 1.60 diff -u -r1.60 swank-allegro.lisp --- swank-allegro.lisp 1 Oct 2004 12:16:44 -0000 1.60 +++ swank-allegro.lisp 24 Oct 2004 18:55:03 -0000 @@ -419,7 +419,7 @@ `("Name: " (:value ,(function-name f)) (:newline) "Its argument list is: " ,(princ-to-string (arglist f)) (:newline) "Documentation:" (:newline) - ,(documentation f 'function)))) + ,(documentation (excl::external-fn_symdef f) 'function)))) (defmethod inspect-for-emacs ((class structure-class) (inspector acl-inspector)) (values "A structure class."

1 0

[slime-devel] (in-package foo)
by Neo-LISPer 23 Oct '04

23 Oct '04

When I SLIME-EVAL-LAST-EXPRESSION something like (IN-PACKAGE FOO), it does not change *PACKAGE*. Any way to fix this?

4 3

[slime-devel] indentation!
by Pascal J.Bourguignon 22 Oct '04

22 Oct '04

How can we disable slime indentation and revert to the normal indentation? -- __Pascal Bourguignon__ http://www.informatimago.com/ Voting Democrat or Republican is like choosing a cabin in the Titanic.

2 2

[slime-devel] Confused Meta-Point
by d95-bli＠nada.kth.se 20 Oct '04

20 Oct '04

I wonder if there is some way to reset the information that MP uses to find definitions. Right now, I have a running image with a file which I have done a lot of editing on, and although MP happily finds definitions, it finds completely wrong ones. For instance I have a generic function where if I MP it, I get the list with the GF itself and all methods, but each one of those when selected jumps to a completely unrelated form in the source. Björn

2 2

[slime-devel] Is this right?
by Peter Seibel 20 Oct '04

20 Oct '04

These days (and perhaps before) function arguments get indented like this: (defun foo (x y z) (some-function x y z)) while I was sort of expecting the x, y, and z to be in one more space like this: (defun foo (x y z) (some-function x y z)) Have I slipped a cog? I can't find any variable in the customizable lisp indentation variables that controls this. Has it always been this way? -Peter -- Peter Seibel peter(a)javamonkey.com Lisp is the red pill. -- John Fraser, comp.lang.lisp

2 1

[slime-devel] Daily ChangeLog diff
by Luke Gorrie 19 Oct '04

19 Oct '04

Index: slime/ChangeLog diff -u slime/ChangeLog:1.552 slime/ChangeLog:1.555 --- slime/ChangeLog:1.552 Sun Oct 17 23:51:40 2004 +++ slime/ChangeLog Tue Oct 19 20:59:52 2004 @@ -1,3 +1,15 @@ +2004-10-19 Luke Gorrie <luke(a)synap.se> + + * slime.el (slime-show-source-location): Call `push-mark' to push + the source position onto the global mark ring. + +2004-10-19 Helmut Eller <heller(a)common-lisp.net> + + * swank.lisp (define-printer-variables): NIL is not a valid + docstring. Reported by Alain Picard. + (printer-variables sldb-print): Include print-gensym, + pprint-dispatch, base, radix, array, and lines. + 2004-10-17 Luke Gorrie <luke(a)synap.se> * slime.el (slime-message): Use slime-typeout-frame if available.

1 0

[slime-devel] Location of read errors
by Gerd Flaig 19 Oct '04

19 Oct '04

Hi, when I compile a buffer using C-c C-k and the source triggers a read error, I get thrown into sldb which shows something like this: Error in function LISP::FOP-EVAL-FOR-EFFECT: Attempt to load a file having a compile-time read error. [Condition of type SIMPLE-ERROR] Restarts: 0: [CONTINUE] Skip this form. 1: [CONTINUE] Return NIL from load of #p"/home/gefla/src/macho-0.4/src/message.x86f". 2: [ABORT] Abort SLIME compilation. 3: [ABORT] Abort handling SLIME request. 4: [ABORT] Return to Top-Level. Backtrace: 0: (LISP::FOP-EVAL-FOR-EFFECT) 1: (LISP::LOAD-GROUP #<Stream for file "/home/gefla/src/macho-0.4/src/message.x86f">) 2: (LISP::FASLOAD #<Stream for file "/home/gefla/src/macho-0.4/src/message.x86f">) 3: (LISP::INTERNAL-LOAD #p"/home/gefla/src/macho-0.4/src/message.x86f" #p"/home/gefla/src/macho-0.4/src/message.x86f" :ERROR :BINARY) 4: (LISP::INTERNAL-LOAD #p"/home/gefla/src/macho-0.4/src/message.x86f" #p"/home/gefla/src/macho-0.4/src/message.x86f" :ERROR NIL) 5: (LOAD #p"/home/gefla/src/macho-0.4/src/message.x86f" :VERBOSE T :PRINT ...) 6: (COMPILE-FILE "/home/gefla/src/macho-0.4/src/message.lisp" :OUTPUT-FILE T :ERROR-FILE ...) 7: ("DEFIMPLEMENTATION SWANK-COMPILE-FILE") [...] .. which is not too helpful. Pressing 'v' to view the source location shows the source for fop-eval-for-effect, which is not exactly where I wanted to go. The real error is buried in the REPL buffer. ; Compiling DEFUN PARSE-MIME-HEADER: ; ; Error: Read error at 7442: ; "(or (net.telent.date:parse-time/\ (cadr (split ";\\s*" received)))" ; Reader error at 7476 on #<Stream for file "/home/gefla/src/macho-0.4/src/message.lisp">: ; package "NET.TELENT.DATE" not found ; Did I misconfigure something or are read errors not handled in a sane way yet? Goodbyte, Gerd. -- Gerd Flaig Technik gerd(a)schlund.de Bei Schlund + Partner AG Brauerstraße 48 D-76135 Karlsruhe Physics is like sex: sure, it may give some practical results, but that's not why we do it. -- Richard Feynman

2 3

[slime-devel] Problem with viewing source with v from debugger
by Peter Seibel 19 Oct '04

19 Oct '04

My problem is not that I can't view source--works fine in Allegro 6.2 on GNU/Linux, fairly recent CVS Emacs, etc. My problem is that after v'ing around a few stack frames to find the code I need to fix I hit q in the sldb buffer to abort and the source I was just looking at disappears jumping me back to who know's where--probably what ever code my cursor was on before I evaluating whatever expression signaled the error. Is there any way to get the source to stay put? -Peter -- Peter Seibel peter(a)javamonkey.com Lisp is the red pill. -- John Fraser, comp.lang.lisp

2 1

[slime-devel] Macro indentation redux
by Peter Seibel 19 Oct '04

19 Oct '04

Okay, I'm totally confused. I have a couple macros with lambda lists like this: (defmacro with-gensyms ((&rest names) &body body) ... (defmacro once-only ((&rest names) &body body) ... It seems that a few weeks ago, when I used these macros Emacs would indent the bodies correctly, i.e. (with-gensyms (foo) (stuff ...)) and (once-only (foo) (stuff ...)) Then something changed and once-only started getting indented like this: (once-only (foo) (stuff ...)) Now (having just updated to CVS head) once-only gets indented correctly but with-gensyms is dorked up: (with-gensyms (foo) (stuff ...)) Help? (BTW, I'm using a fairly recent CVS version of Emacs--a week or two old, FWTW.) -Peter -- Peter Seibel peter(a)javamonkey.com Lisp is the red pill. -- John Fraser, comp.lang.lisp

5 13

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

slime-devel