- pro - mailman3.common-lisp.net

[pro] Fwd: Re: Lisp and DSLs
by raito＠raito.com 20 Jul '11

20 Jul '11

This may not be much of an example, but it's not only what I have, it's what got me hooked on Lisp. When I was a young lad, and because I lived not too far from Lake Geneva, I played this newfangled thing called D&D (back when it first came out. Get off my lawn). And one of the things that intrigued me was the tables (the kind where you rolled dice and looked up the results). And also being as I was learning programming (on a mainframe. There were no microcompputers. See my previous comment about lawns.) I thought about writing a table language. One where I could specify not only how many chances a particular result had (because I'd already figured out that that was better than using the dice themselves for a bell curve), but also being able to modify the tables themselves at run-time by feeding them arguments, being able to have sub-tables, detailed results, etc. But at the time all we had was Honeywell mainframe BASIC (supposedly FORTRAN was available, but we couldn't get a manual), and even the instructor didn't know anything about writing parsers, let alone languages. Fast forward, and the idea stuck with me. I mean, I'd gone to school and doubles in ECE and Comp. Sci., I should be able to do this. But I never did, because it's a very big hassle to write your own language. Especially in languages like C. So when I started using Lisp, I read up all this stuff about DSLs. Hm, I wonder if I could use Lisp to do my language. Well, half an hour and 15 lines of code later, I not only had my language, but more than I wanted. And here it is (ta da): (defmacro table (name args &rest rest) `(defun ,name (,@args) (declare (special ,@args)) (let* ((base ',(loop for x in rest collect (car x))) (results ',(loop for x in rest collect (cadr x))) (chances (loop for x in base collect (eval x))) (total (loop for x in chances sum x)) (roll (+ 1 (random total)))) ;(format t "Base:~{~a~^, ~}~%Chances:~{~a~^, ~}~%Results:~{~a~^, ~}~%total:~A roll:~A~%" ; base chances results total roll) (do ((n chances (cdr n)) (cnt 0) (item 0 (+ item 1))) ((eq n nil) nil) (setf cnt (+ cnt (car n))) (when (<= roll cnt) (return (eval (nth item results)))))))) and it's used like this (a simple example): (table my-table (added-chances) (1 "one") (1 "two") ((+ 1 added-chances) "three") ) I've done some stuff with it that's been quite a bit of fun. Like another layer of macro on top for those who think in terms of dice. The upshot is that it's a project I didn't do for 25 years because it was too big a pain. And I did it in Lisp in half an hour, including figuring out that I had to use dynamic scoping for args. I have to use dynamic scoping because the forms being eval'ed need to be able to access args, but args is not in their lexical scope (though it looks as if they are in the macro invocation -- which is the point). Neil Gilmore raito(a)raito.com

1 0

[pro] generating documentation
by Tamas Papp 19 Jul '11

19 Jul '11

Hi, I am about to complete a revision of LLA and I thought I would write some decent documentation. I have most of the documentation in docstrings, though some of general discussions are in ;;; comments in files. I am looking for a reasonably automatic way of generating documentation (eg texinfo or HTML) from the source. I found many libraries for this on cliki, and instead of trying all of them out, I thought I would ask for advice here. Portability would be nice is not necessarily a requirement, so tools like sbcl's texinfo are fine too. However, if possible, I would like to put the documentation on the interface and the internal details in separate sections somehow. I would also be interested in including tutorial-like sections with examples. Thanks, Tamas PS.: I have searched the cl-pro and c.l.l archives and found some discussions about this question, but my impression was that a lot of new libraries appeared on the scene in the past few years.

3 3

[pro] Is gensym used for anything except alpha-conversion in macros?
by Kazimir Majorinc 11 Jul '11

11 Jul '11

Is gensym used for anything except alpha-conversion in macros?

5 9

Re: [pro] Is gensym used for anything except alpha-conversion in macros?
by Steve Haflich 10 Jul '11

10 Jul '11

On Jul 10, 2011 2:09 AM, "David McClain" <dbm(a)refined-audiometrics.com> wrote: > > Gensym can also be used to produce globally (session wide) unique objects. But this is no different from other objects, such as a cons or array. The primary difference between an uninterned symbol and any other object that has settable "slots" is its behavior as a form under evaluation. Arrays and most other objects self evaluate; a cons has complex interpretation as a form; but a symbol evaluates to its its current special or lexical value binding. That's why gensyms are mostly use as hygienic names in macro expansions. There are other ways of doing whar a ge.sym does. Maybe later when I'm not typing on a tiny phone keyboad i'll show one.

1 0

[pro] LISP technology at Cisco
by Daniel Weinreb 08 Jul '11

08 Jul '11

This is a video about "LISP Technology at Cisco". http://www.youtube.com/watch?v=JIiBMZ86jZo&feature=player_embedded (It has nothing to do with Lisp, either.) :) -- Dan

4 3

[pro] why :key arguments?
by Tamas Papp 05 Jul '11

05 Jul '11

Why do some CL library functions have :key arguments? I am asking because I am working on some statistics functions, and the design choice came up. Specifically, I can write functions like (defun quantiles (sequence quantiles &key (key #'identity)) ...) but it is a bit cumbersome. I can make my code simpler by relying on calls like (quantiles (map 'vector key vector) quantiles) but this conses a bit more. Not a major concern at the moment, but if it ever becomes one, I could just define a compiler macro to take care of forms like this and transform to something like (quantile-with-keys vector quantiles key) I also thought of defining something like (defgeneric map1 (function object) (:method (function (list list)) (mapcar function list)) ...) for the sole purpose of mapping objects to similar objects (eg lists to lists, arrays to arrays, etc) and allow it to be optimized away (like above) by compiler macros. I would appreciate advice on this. I am especially interested in the reason why some CL functions have :key arguments: is it because of efficiency, backward-compatibility/history, or something else? Thanks, Tamas

10 20

[pro] Generated symbols
by Kazimir Majorinc 04 Jul '11

04 Jul '11

I need comprehensive review of use of generated symbols (make-symbol, copy-symbol, gensym... ) in CL. The books I have (both Grahams, Siebel, barely mention these operators, except for use of gensym in macros. What are other typical or interesting uses of these? -- Kazimir Majorinc

3 2

[pro] "fhash"
by Daniel Weinreb 30 Jun '11

30 Jun '11

Friends, I wrote a little package for "fash hash tables", which provide an abstraction that is analogous to that of Common Lisp hash tables, but is faster for tables with few elements, and only slightly inferior for tables with many elements. I did this because performance analysis showed that our system was spending too much time in hash table operations, and using the new package helped. I have recently been cleaning this up, one reason being that I'd like to open source it. The function names used to be things like getfhash and mapfhash. Now they are like fhash:get and fhash:map-elements and so on. However, before I open-source it, I was to make sure it's "right". It recently occurred to me that the package name "fhash" has problems. Here are pros and cons of changing it that I can see. Pro: I's not a hash table in the small-cardinality case; it's a linear lookup. So the name is not actually accurate. Pro: Calling such a data structure a "hash table", even as Common Lisp does, is an abstraction violation. Whether it works by hashing is an implementation detail. The Java collection library calls this a Map. Python calls it a dictionary. Clojure calls it a map. Those are both better names. Con: Common Lisp already uses the name "hash table", so it would be easier for existing Common Lisp programmers to see the analogy. Advice? Thanks! Thanks! -- Dan

8 13

[pro] library configuration
by Hans Hübner 29 Jun '11

29 Jun '11

I made a moderation error, so I'm forwarding this message manually. -Hans ---------- Forwarded message ---------- From: Tamas Papp <tkpapp(a)gmail.com> To: pro(a)common-lisp.net Date: Tue, 28 Jun 2011 12:26:24 +0000 (UTC) Subject: library configuration Hi, A user of my LLA library asked for more flexibility in specifying which foreign library to load. For example, BLAS/LAPACK come in so many variants that it is impossible to cover them with the :or feature of cffi:define-foreign-library, especially because users have their own preferences when more than one is available. I am looking for a portable way to specify configuration for a library before/during loading with ASDF2. It would be sufficient if the user could specify an ALIST (or something similar) somewhere and I could read/ use it when the library is loaded. Does ASDF2 have a "standard" mechanism for library config files (eg where they are, how to handle them etc), or should I use something else? I thought of using *features*, but I need actual values instead of merely checking for the presence of symbols. Any suggestions are welcome. Thanks, Tamas

6 8

[pro] Some error/condition standardization anyone?
by Marco Antoniotti 28 Jun '11

28 Jun '11

Hi I have the following problem (here below in CCL and LW) CCL: marcoxa$ ccl AM I BEING LOADED? Welcome to Clozure Common Lisp Version 1.5-r13651 (DarwinX8632)! ? (ignore-errors (read)) foo::bar NIL #<CCL::SIMPLE-READER-ERROR #xC48489E> ? LWM: CL-USER 1 > (ignore-errors (read)) foo: NIL #<CONDITIONS:PACKAGE-NOT-FOUND-READER 21BA3B93> CL-USER 2 > Other Lisps behave similarly. I.e. they don't agree on the type of the error that they need to generate. Now, this is a specific case for me: having a consistent READER-ERROR-BECAUSE-THERE-IS-NO-SUCH-PACKAGE-THAT-THE-SYSTEM-KNOWS error signaled would simplify some code writing. Doesn't anybody think that it would be a good thing to have a CDR suggesting an agreed upon extension of the CL standard ERROR/CONDITION hierarchy? Cheers -- Marco Antoniotti, Associate Professor tel. +39 - 02 64 48 79 01 DISCo, Università Milano Bicocca U14 2043 http://bimib.disco.unimib.it Viale Sarca 336 I-20126 Milan (MI) ITALY Please note that I am not checking my Spam-box anymore. Please do not forward this email without asking me first.

5 5