Re: Announce: PicoLisp in Hardware (PilMCU)
Oh boy! I've been thinking of something like this for a while.
What is the low-hanging fruit here, in the sense of, what applications might we
do faster/cheaper/better in hardware like this, than can be done otherwise?
The idea of pilog in hardware excites me too ... maybe time to go back and dust
off old 1980s Fifth Generation Computer Systems project application ideas -
http://en.wikipedia.org/wiki/Fifth_generation_computer
Maybe some cool application that can be done better now with pilog in hardware
than could be done otherwise. Same for the old lisp-machine/Symbolics
applications, I wonder if there is anything useful from Symbolics' experience,
in retrospect?
The Fifth Generation Computer project was considered a failure but seems to be
relevant to to picolisp (and pilog) in hardware.
It would be really cool if picolisp in hardware ends up accomplishing things
that caused Symbolics Inc and The Fifth Generation Computer Systems project to
stumble. I expect this to be the case.
Doug
On Fri, 9/19/14, Alexander Burger a...@software-lab.de wrote:
Subject: Announce: PicoLisp in Hardware (PilMCU)
To: picolisp@software-lab.de
Date: Friday, September 19, 2014, 4:39 AM
Hello List,
we are proud to announce PilMCU, the Lisp Machine on a Chip!
:)
We, that is George Orais (who persuaded me into the project)
and me.
Georg built the actual machine in Verilog, and I did the
changes and
extensions to PicoLisp.
PilMCU is an implementation of 64-bit PicoLisp directly in
hardware. A
truly minimalistic system. PicoLisp is both the machine
language and the
operating system:
* Memory management is trivial, just the
Lisp heap and the stack
* The built-in database is extended to
hold a file system
* One SSD per database file for mass
storage
* Processes run as tasks and coroutines
* Events (timing and interrupts) via a
'wait' instruction
* Complex I/O protocols are delegated to
peripheral chips
The final hardware can be very lightweight. Low transistor
count and
power consumption. No overhead for an OS. It is conceivable
for a later
stage to put many interconnected CPUs on a single chip.
At present, we have it running in the Verilog simulator, and
in an
emulator (adaption of the PicoLisp 'emu' architecture).
How shall we proceed? We need investors (or crowdfunding) to
polish,
manufacture and distribute the real thing.
We imagine something in the line of an Embedded Lisp
Machine or a
Lisp Machine Kit. Perhaps for home brewing, educational
institutions
and/or robotics research?
Is anybody interested -- or knows people who are?
For the fun of it, here is a sample session:
$ make mcu vvp -M.
-mtty mcu # Build and start Verilog engine
:
$ make emu ./emu
ssd@ ssdA # Or: Build and start the
emulator
:
Now we are in an environment equivalent to the standard 'pil
+'. The
database is open on two image files for two SSD drives.
Besides the
normal, full DB functionality
: (show *DB)
{1} (7 . {17})
T ({2} {20} {56} {64} {105} {146})
- {1}
you can call 'in', 'out', 'load' and 'rm' on files which are
maintained
in external symbols:
: (dir)
- (lib.l lib/)
: (dir lib)
- (btree.l db.l dbg.l misc.l
pilog.l sq.l)
: (in lib/db.l (read))
- (de dbs Lst (default *Dbs (_dbs
1)))
: (out foo/bar/mumble.l (prinl Hello
world))
- Hello world
: (in foo/bar/mumble.l (line))
- (H e l l o w o r
l d)
: (dir foo/bar)
- (mumble.l)
: (cd foo/bar)
- foo/bar/
: (dir)
- (mumble.l)
: (pwd)
- foo/bar/
Path names are stored as a normal B-Tree in the DB root:
: (scan)
foo/bar/mumble.l {172}
lib.l {2}
lib/btree.l {64}
lib/db.l {105}
lib/dbg.l {20}
lib/misc.l {56}
lib/pilog.l {146}
lib/sq.l {166}
They point to external symbols, like {2} for lib.l.
(load '{2})
is equivalent to
(load lib.l)
The values of these symbols hold the file size:
: (show '{2})
{2} 12401
- {2}
They should not have properties, and store the raw file data
invisibly
in dynamically maintained DB blocks.
The rest of the system is standard PicoLisp :)
♪♫ Alex
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: Android Demo
I'll have to review my notes and see where I was on that. After some initial successes compiling, after testing more, I remember some snags in either mini picolisp or full picolisp involving some ARM vs x86 differences. --- On Tue, 10/9/12, Thorsten Jolitz tjol...@googlemail.com wrote: From: Thorsten Jolitz tjol...@googlemail.com Subject: Re: Android Demo To: picolisp@software-lab.de Date: Tuesday, October 9, 2012, 10:06 AM Doug Snead semaphore_2...@yahoo.com writes: Hi Joe and Doug, I read up your conversations about PicoLisp on Android in the mailing list and your articles in the Wiki about the topic: Ok, so now this means we can package some picolisp client/server applications as android apps, using the android browser. (I think.) I like that approach because it works with the android browser and app packaging scheme - and at the same time you're getting picolisp in all its glory and efficiency. I don't have an Android device to try out your code examples, but I would be interested in the 'status quo' of this topic - is full PicoLisp running on Android, have there been any real Apps implemented in PicoLisp by somebody? If so - are there any cons of using PicoLisp instead of Java to be considered? -- cheers, Thorsten -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog in ersatz
The pilog trace can be helpful in these situations. : (be bigger (me her)) : (be bigger (her son)) : (be bigger (son daughter)) : (be bigger (@x @y) (bigger @x @z) (bigger @z @y)) List the assertions you want to trace before the clause to be proved, in this case bigger : : (? bigger (bigger @x meily)) 4 (bigger @x meily) 1 (bigger me her) 4 (bigger her meily) 2 (bigger her son) 4 (bigger son meily) 3 (bigger son daughter) 4 (bigger daughter meily) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) 4 (bigger daughter @y) ... So, it is infinitely recurring when trying to prove (bigger daughter @y). The numbers in front of the pilog trace output lines correspond to the rule being proved at that stage. Hope that helps! Cheers, Doug --- On Sun, 7/8/12, Christophe Gragnic christophegrag...@gmail.com wrote: From: Christophe Gragnic christophegrag...@gmail.com Subject: pilog in ersatz To: picolisp@software-lab.de Date: Sunday, July 8, 2012, 11:36 AM Hi all, I'm currently investigating Picolisp to teach prefix notation, and Pilog to teach first order logic. I have a problem with a set of instructions in ersatz: : (be bigger (me her)) - bigger : (be bigger (her son)) - bigger : (be bigger (son daughter)) - bigger : (be bigger (@x @y) (bigger @x @z) (bigger @z @y)) - bigger : bigger - NIL : (? (bigger @x meily)) @x=alix @x=sw Then ersatz freezes, no prompt. Any idea? chri -- Envoyé de ma messagerie électronique. Merci de divulguer cette adresse mail au minimum (pour les envois en groupe, utilisez la copie invisible). -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog in ersatz
--- On Sun, 7/8/12, Christophe Gragnic christophegrag...@gmail.com wrote: : (be bigger (me her)) - bigger : (be bigger (her son)) - bigger : (be bigger (son daughter)) - bigger : (be bigger (@x @y) (bigger @x @z) (bigger @z @y)) - bigger : bigger - NIL : (? (bigger @x meily)) @x=alix @x=sw Hmmm. Where do alix and sw originate? I do not understand that. (I was using full picolisp for my tracing before, I should have mentioned.) -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: Android Web Server
Hi Joe, Very nice! I think it would be useful to be able to package complete picolisp applications as android apps. Or, at least that might be a possibility. Something to explore. I'm still hung up on trying to get android browser to talk to the picolisp app server like that. I'm reading there are some peculiarities with android localhost which is what I think I am running into now. Also I am thinking about best ways to have android java interface with pilog too. Cheers, Doug --- On Thu, 2/23/12, Joe Bogner joebog...@gmail.com wrote: From: Joe Bogner joebog...@gmail.com Subject: Android Web Server To: picolisp@software-lab.de Date: Thursday, February 23, 2012, 7:42 AM Doug Snead's recent posts have encouraged me to write up my progress along similar lines with getting PicoLisp to work on Android. http://picolisp.com/5000/!wiki?AndroidWebServer I have a functioning web server. My approach was to cross compile on unix and then copy the binary to the device. If anyone has any input on the issue at the end I would appreciate it. I also uploaded the code in case I decide to remove it off my public dropbox. Thanks and I look forward to any comments! Joe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: server - IP socket error: Address family not supported by protocol
Might be even easier to reproduce ... just try the port function: # picoLisp-2.2.7.tgz ... ~/lisp/picoLisp $ ./p : (port 1234) - 3 That's ok, but the most recent version, # picoLisp-3.0.9.tgz # ./pil : (port 1234) !? (port 1234) IP socket error: Address family not supported by protocol ? --- On Wed, 2/22/12, Doug Snead semaphore_2...@yahoo.com wrote: From: Doug Snead semaphore_2...@yahoo.com Subject: server - IP socket error: Address family not supported by protocol To: picolisp@software-lab.de Date: Wednesday, February 22, 2012, 10:07 PM Hi Alex, I'm tying to get the example in doc/app.html (picoLisp-3.0.9.tgz on a pretty standard linux box) to serve. Here's my .l file: # cat test_html_app.l (html 0 Hello @lib.css NIL Hello World! ) (renamed from project.l but otherwise identical) But when I try to start the server, like this: # ./pil @lib/http.l @lib/xhtml.l @lib/form.l -'server 8080 test_html_app.l' + Picolisp tells me: !? (port *Port) IP socket error: Address family not supported by protocol ? Same for a different port number: # ./pil @lib/http.l @lib/xhtml.l @lib/form.l -'server 54321 test_html_app.l' + !? (port *Port) IP socket error: Address family not supported by protocol ? This is the linux platform I am using (Linux 2.6.30.5 #1 SMP Tue Sep 1 15:48:26 GMT-8 2009 i686 GNU/Linux). I could easily be doing something wrong ... but I think the above should work. Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: server - IP socket error: Address family not supported by
Hi Alex, Ah, ok ... hmmm, that's two different platforms I have seen that problem on, then. The Android NDK also compiles the latest picoLisp ok, but then, same as that other (more standard) linux, dies in that port function with the same error. I can drop back to a pre-IPv6 version of picolisp and try that. Cheers, Doug --- On Wed, 2/22/12, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: server - IP socket error: Address family not supported by To: picolisp@software-lab.de Date: Wednesday, February 22, 2012, 10:53 PM Hi Doug, IP socket error: Address family not supported by protocol This error usually has a very simple reason: Your platform doesn't support IPv6 Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
android success with full picolisp
More android + picolisp fun, this time with the full picolisp. Using the android SDK and NDK, I hacked a picolisp/src/makefile to work for android's arm processor like this: --- makefile --- [snip] CFLAGS := -c -O2 -pipe \ -falign-functions=64 -fomit-frame-pointer -fno-strict-aliasing \ -W -Wimplicit -Wreturn-type -Wunused -Wformat \ -Wuninitialized -Wstrict-prototypes \ -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 # ?? had: -m32 NDK_ROOT = ~/android/android-ndk-r7 NDK_BIN = $(NDK_ROOT)/toolchains/arm-linux-androideabi-4.4.3/prebuilt/linux-x86/bin SYS_ROOT = $(NDK_ROOT)/platforms/android-8/arch-arm/ CC = $(NDK_BIN)/arm-linux-androideabi-gcc --sysroot=$(SYS_ROOT) LD = $(NDK_BIN)/arm-linux-androideabi-ld AR = $(NDK_BIN)/arm-linux-androideabi-ar RANLIB = $(NDK_BIN)/arm-linux-androideabi-ranlib STRIP = $(NDK_BIN)/arm-linux-androideabi-strip OS = Arm PICOLISP-FLAGS = -m32 -rdynamic LIB-FLAGS = -lc -lm -ldl DYNAMIC-LIB-FLAGS = -m32 -shared -export-dynamic [snip] - Then (to my surprise) picolisp and dynamic libraries were made, # file ../bin/picolisp ../lib/ext ../lib/ht ../lib/z3d ../bin/picolisp: ELF 32-bit LSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), stripped ../lib/ext: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), stripped ../lib/ht: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), stripped ../lib/z3d: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), stripped So far so good... Using this android approach generally, http://gimite.net/en/index.php?Run%20native%20executable%20in%20Android%20App I placed the picolisp executable in the assets dir and at run-time, copy it from assets to /data/data/ in the right place for that app. Since I'm using the emulator and I know where the executable was placed, I can run it using adb, for some command-line tests: # adb shell /data/data/com.mytest/picolisp '-de foo (X) (println X)' '-foo 123' -bye 123 # adb shell /data/data/com.mytest/picolisp '-de foo (X) (println (* X 2))' '-foo 123' -bye 246 # adb shell /data/data/com.mytest/picolisp '-de foo (X) (println (* X 2))' '-foo 12345' -bye 24690 A bit cumbersome having to unpack the executable and other files from the app's .apk (zip archive) to run it ... but it can be done. And no fiddling with bits ... no changes to the (full picolisp) source at all. Next step is to try to similarly unpack all the libraries and see if a picolisp database server application can be run. Then more testing. And use that with android's browser, all within an android app. But I'm very confident that the full picolisp will run on the android from what I see so far! There are ways to call java from C also, so that opens up possibilities of using android java libraries from android picolisp too. Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: miniPicoLisp Transients
--- On Thu, 1/26/12, Alexander Burger a...@software-lab.de wrote: (setq V 456) at the top level (not in a definition) that works as expected Again strange. It shouldn't matter how the symbol is used (globally or bound locally). Hi Alex, Thanks for the guidance! On second look, I was wrong about (setq V 456) ... it returns V from the top level on the (incomplete) android arm port. Close, though. :-) Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
miniPicoLisp Trasnients
Alex, I have a porting-related question for you. After more android testing I'm finding a problem with (transient) variables ... (I think). When I try a definition like this, (de foo (x) x) no problem, (foo 123) returns 123 as expected. But when I try it this way: (de baz (X) X) (baz 123) I should get 123 but instead I get: X However, if I try (setq V 456) at the top level (not in a definition) that works as expected, and evaluating V later returns 456 I wonder why it might not see X as a transient? (Er, I'm not even sure if transient is the the right terminology here.) Does anything leap out at you there? :-) Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: PicoLisp is DEAD (Was: PicoLisp and its (lack of) libraries)
After many merry rounds of compilation and discovery, it seems that the ndk-build's compiler -falign-functions[=n] and friends align functions relative to this option, but off by one (+1, or is that just |1?).;-) In pico.h, sadly, I besmirched the code thusly, #ifdef ANDROID_ARM // android arm seems to have weird off-by-one alignment #define FUNCT2LISP_ALIGN(f) ((f)-1) // hopefully optimise to #define LISP2FUNCT_ALIGN(f) ((f)+1) // a single inc ... #else #define FUNCT2LISP_ALIGN(f) (f) #define LISP2FUNCT_ALIGN(f) (f) #endif #define evSubr(f,x) (*(fun)( LISP2FUNCT_ALIGN( num(f) ) ~2 ) )(x) so in boxSubr in main.c, if (num(FUNCT2LISP_ALIGN(f)) 3) giveup(Unaligned Function); return (any)(num(FUNCT2LISP_ALIGN(f)) | 2); (Er, I bet I missed some places, need to check.) And then to talk with java I used the miniPicoLisp string-based library mods mentioned earlier, with a little jni C glue to hold it together. Cheers, Doug --- On Mon, 1/23/12, Jakob Eriksson ja...@aurorasystems.eu wrote: From: Jakob Eriksson ja...@aurorasystems.eu Subject: Re: PicoLisp is DEAD (Was: PicoLisp and its (lack of) libraries) To: picolisp@software-lab.de Date: Monday, January 23, 2012, 11:47 PM On January 24, 2012 at 7:49 AM Doug Snead semaphore_2...@yahoo.com wrote: I have a (slightly hacked) version of miniPicoLisp running as an android native library as a proof of concept. No no additional java interpretation penalty. Wow, this is interesting! My thought is that now I have a new tool - a way to make picolisp and pilog (prolog) work in android, but will do the UI in java like most android apps. That is how I plan to use (mini)PicoLisp in my program too. But miniPicoLisp is pure, and easier to port. Indeed, and for me the most interesting right now. best regards, Jakob -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: PicoLisp is DEAD (Was: PicoLisp and its (lack of) libraries)
--- On Mon, 1/23/12, Jakob Eriksson ja...@aurorasystems.eu wrote: The Java version opens up the path to Android jobs - and I have an impression that those are still a kind of Wild West. I have a (slightly hacked) version of miniPicoLisp running as an android native library as a proof of concept. No no additional java interpretation penalty. The C-compiler (gcc?) for the ARM processor used in android ndk-build has odd ideas about function alignment that gave me a hard time until I saw what was happening. My thought is that now I have a new tool - a way to make picolisp and pilog (prolog) work in android, but will do the UI in java like most android apps. Not sure if porting the full PicoLisp to androig ndk jni would be easier or harder then the Cygwin port, say. The Cygwin port wasn't difficult ... until we discovered the Many Joys of Win32 file locking. But miniPicoLisp is pure, and easier to port. :-) -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
re: tags
FWIW ... here's something I use instead of ctags for picolisp.
$ cat /usr/local/bin/lisptags
#!/bin/sh
# make a tags file for pico lisp source files.
# use:
# lisptags foo.l bar.l baz.l ... bof.l
# generate the file 'tags'
# [based on lisptags csh script by John Foderaro, c.1982]
awk '
/^[ \t]*\(d[em][ \t]/ { print $2 \t FILENAME \t?^ $0 $? }
/^[ \t]*\(class[ \t]/ { print $2 \t FILENAME \t?^ $0 $? }
' $* | sort tags
Doug
--- On Mon, 11/14/11, Henrik Sarvell hsarv...@gmail.com wrote:
From: Henrik Sarvell hsarv...@gmail.com
Subject: Re: How to get the signature of function and method definitions
To: picolisp@software-lab.de
Date: Monday, November 14, 2011, 6:59 AM
It's at the end here:
http://www.prodevtips.com/2010/09/29/emacs-color-themes-tags-cedet-ecb-and-other-customizations/
Not much to it, I do things old school ie I put the
download in
/opt/picolisp and simply cd there in a shell and do: ctags
-e -R
--languages=-JavaScript,-PHP,-C,-Make,-HTML
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: error message line numbers
Thanks for the reply, Alex. I need to use Windows without cygwin dlls for an application I'm working on now. So, I had been using miniPicoLisp in my application, which compiles ok under mingw. That let me avoid even trying to compile regular PicoLisp using mingw, and the porting issues that would involve. But now I might have to take a look at trying PicoLisp with mingw. I know the file locking will probably be an issue, but not sure what else might be a snag with PicoLisp under mingw. On the other hand, I was able to hack in some support for line numbers and file names for error messages in miniPicoLisp, relatively quickly. Cheers, Doug --- On Sun, 9/4/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: error message line numbers To: picolisp@software-lab.de Date: Sunday, September 4, 2011, 11:29 PM Hi Doug, Is there a way I can make (mini)PicoLisp give me line numbers from .l source files in error messages? There are lots of gnu and/or linux tools (like vim) and IDEs etc., that want error messages in a format similar to what gcc spits out, like: Baz.cpp:321: error: 'foobar' was not declared in this scope (I poked around in the code and on the site looking, but missed it if it is there.) All other PicoLisp implementations (64-bits, 32-bits and Ersatz) have this feature. They keep track of the file name and line number in the 'inFile' structure (or LineNumberReader in case of Java). For MiniPicoLisp this is difficult, because it uses plain stdio instead of a dedicated 'inFile' structure. Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
error message line numbers
Hi Alex, Is there a way I can make (mini)PicoLisp give me line numbers from .l source files in error messages? There are lots of gnu and/or linux tools (like vim) and IDEs etc., that want error messages in a format similar to what gcc spits out, like: Baz.cpp:321: error: 'foobar' was not declared in this scope (I poked around in the code and on the site looking, but missed it if it is there.) Cheers Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog dcg with args
Hi Alex, Thanks for the picolisp pointers :-) The sub? shrinks down one of the awkward bits there. Unfortunately, I didn't see sub? in miniPicoLisp. The replace seems like a better idea than using the surgical patch there - probably not a good idea to modify a list passed in as an argument like I am doing with patch. Cheers, Doug --- On Sun, 7/24/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: pilog dcg with args To: picolisp@software-lab.de Date: Sunday, July 24, 2011, 2:36 AM Hi Doug, http://victoriafamilyrobotics.net/vfr1/media/read_line.l Please allow me to post some minor hints: When a single character is checked for a number of values like in (or (= X ,) (= X .) (= X ?)) then I would write instead: (sub? X ,.?) For the modification of a list (patch L , '( ,)) (patch L . '( .)) (patch L ? '( ?)) I would use (replace L , '( ,) . '( .) ? '( ?)) Note, however, that 'replace' is non-destructive, so you need to use the returned value in the following expression (or 'setq' it to 'L'). Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog and Definite Clause Grammar (DCG)
THANK YOU! I haven't tested that yet but I think that is exactly what I was looking for there. Might be something to consider adding to pilog.l Seems to be fairly standard in prologs - and useful for creating parsers :-) Beautiful code, thanks again so much! Cheers, Doug --- On Wed, 7/20/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: pilog and Definite Clause Grammar (DCG) To: picolisp@software-lab.de Date: Wednesday, July 20, 2011, 2:20 AM Hi Doug, What's the best way to handle Definite Clause Grammar syntax in pilog? .. So, I'm hoping for some pilog rules or a macro maybe that can handle the -- (or equivalent) of DCG in pilog :-) I don't fully understand the context or purpose, but http://en.wikipedia.org/wiki/Definite_clause_grammar says that DCG notation is just syntactic sugar for normal definite clauses in Prolog. So you can write a function 'dcg' which does the necessary translation. I would simply call 'be' here, using 'macro' because 'be' doesn't evaluate its arguments (i.e. is a FEXPR): (de dcg CL (macro (be ^(if (pair (cadr CL)) (list (car CL) (list (cons (caadr CL) '@A) '@A)) (cons (car CL) '(@In @Out) (let Vars (make (link @In) (for (I . @) (cddr CL) (link (pack @ I)) ) (link @Out) ) (mapcar list (cdr CL) Vars (cdr Vars) ) ) ) ) ) ) ) With that, the example in Wikipedia could be written as (dcg sentence nounPhrase verbPhrase) (dcg nounPhrase det noun) (dcg verbPhrase verb nounPhrase) (dcg det [the]) (dcg det [a]) (dcg noun [cat]) (dcg noun [bat]) (dcg verb [eats]) This expands to : (rules 'sentence 'nounPhrase 'verbPhrase 'det 'noun 'verb) 1 (be sentence (@In @Out) (nounPhrase @In @1) (verbPhrase @1 @Out)) 1 (be nounPhrase (@In @Out) (det @In @1) (noun @1 @Out)) 1 (be verbPhrase (@In @Out) (verb @In @1) (nounPhrase @1 @Out)) 1 (be det ((the . @A) @A)) 2 (be det ((a . @A) @A)) 1 (be noun ((cat . @A) @A)) 2 (be noun ((bat . @A) @A)) 1 (be verb ((eats . @A) @A)) - verb Does this help? Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
pilog and Definite Clause Grammar (DCG)
Alex (or anyone), What's the best way to handle Definite Clause Grammar syntax in pilog? Definite Clause Grammar (DCG) and Prolog supports a special rule syntax for writing DCGs. The syntax is simpler, much closer to the syntax one uses in writing context-free grammar rules. When using the DCG syntax, the programmer doesn't have to write all the string variables threaded through the nonterminal procedure calls... http://www.cs.sunysb.edu/~warren/xsbbook/node10.html http://www.csupomona.edu/~jrfisher/www/prolog_tutorial/7_1.html etc. So, I'm hoping for some pilog rules or a macro maybe that can handle the -- (or equivalent) of DCG in pilog :-) Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
pilog: Prolog Tutorial, 2.19 Actions and plans
Here's another prolog-to-pilog translation, this time from John Fisher's excellent Prolog Tutorial, 2.19 Actions and plans http://www.csupomona.edu/~jrfisher/www/prolog_tutorial/2_19.html prolog: http://www.csupomona.edu/~jrfisher/www/prolog_tutorial/2_19pl.txt to pilog: http://victoriafamilyrobotics.net/vfr1/media/sec2_19.l This was an easier example than golog.l, but I definitely needed some of the call-related tricks I learned before! To run it, for example, here's do([on(c,b),on(b,a),on(a,table)]) -- shown about three-fourths of the way down the tutorial page -- which is defined as test t_4 in the pilog source file sec2_19.l # ~/lisp/miniPicoLisp $ ./pil sec2_19.l -t_4 -bye 1 (do ((on c b) (on b a) (on a table))) 1 (do_all ((on c b) (on b a) (on a table)) ((on c b) (on b a) (on a table))) 2 (do_all ((on c b) (on b a) (on a table)) ((on c b) (on b a) (on a table))) 1 (achieve (on c b)) 1 (do_all ((on c b) (on b a) (on a table)) ((on c b) (on b a) (on a table))) 1 (do_all ((on b a) (on a table)) ((on c b) (on b a) (on a table))) 2 (do_all ((on b a) (on a table)) ((on c b) (on b a) (on a table))) 1 (achieve (on b a)) 1 (do_all ((on c b) (on b a) (on a table)) ((on c b) (on b a) (on a table))) 2 (do_all ((on c b) (on b a) (on a table)) ((on c b) (on b a) (on a table))) 1 (achieve (on c b)) 1 (do_all ((on c b) (on b a) (on a table)) ((on c b) (on b a) (on a table))) 1 (do_all ((on b a) (on a table)) ((on c b) (on b a) (on a table))) 1 (do_all ((on a table)) ((on c b) (on b a) (on a table))) 3 (do_all NIL ((on c b) (on b a) (on a table))) on/move: 1 (be on (a table)) 2 (be on (b a)) 3 (be on (c b)) 1 (be move (a b table)) 2 (be move (b c table)) 3 (be move (c table b)) 4 (be move (c b table)) 5 (be move (b table a)) 6 (be move (c table b)) Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: The PicoLisp Ticker
I think you created a bot-trap :-) --- On Fri, 7/15/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: The PicoLisp Ticker To: picolisp@software-lab.de Date: Friday, July 15, 2011, 12:58 AM Hi all, let me share my funny experiences about the PicoLisp Ticker, and how it is handled by Mr. Google: http://picolisp.com/5000/-2-1i.html Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
re: Ninety-Nine Lisp Problems: pilog
re: the 24jul10 comment: The more of the higher-numbered problems... are really typical Prolog problems! All of them involve recursive searches in some solution space. http://picolisp.com/5000/-2-1K.html Then perhaps picolisp's Prolog (pilog) would be ideal? I'm always happy to see pilog examples. (BTW - Many people might not realise the heart of pilog, the prove function, is written in C and seems very fast. For some prolog applications, it may be fast or faster than compiled prologs - would like to see some benchmarks to investigate that feeling. But for years, I was under the false impression that pilog was completely written in lisp and so pilog programs were doubly interpreted, but I was wrong: the heart of the pilog interpreter is C.) I'm searching for a pilog A* (Astar) implementation example these days. :-] Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
heap miniPicoLisp question
~/build/miniPicoLisp # ./pil : (heap) - 1 : (heap T) - 0 : Shouldn't I see something else - or is that correct for miniPicoLisp ? Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
android picolisp using ndk/jni
http://developer.android.com/sdk/ndk/overview.html : Typical good candidates for the NDK are self-contained, CPU-intensive operations that don't allocate much memory, such as signal processing, physics simulation, and so on. Ok, here's a plan for :-) a real Android picolisp ... start by taking a version of picolisp, and begin cutting it back by dropping everything ndk c/c++ headers do not support.So, the idea would be you'd get much of picolisp, hopefully more than minipicolisp, at least as much as ersatz, (I would think), and with no java interpretation penalty for the lisp interpreter. re: ... that don't allocate much memory. Fine; allocate what you need per application on init (from java), pass it to android picolisp to use. -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog question
Hi Alex, Ah, the situation is not so bad IMO. To begin with, this is an obscure thing to do, I think - maybe somewhat pedantic. It forces golog to be prolog, in a sense. And there are workarounds which you provided. And I think you're right: it should be possible to restructure the involved golog.l rules to do this better. Maybe another rule or a lisp function can be written, which gets the bindings instantiated and copied out to another, less partially unified, expression. Or maybe a way to make that pilog magic number to bind at the desired level, *and all levels above that also*, perhaps if a negative magic numbers appears there. (But I wouldn't touch doProve, if it slows it down!) In any event, the golog.l source and simple test program are now in the public domain and so hopefully better pilog programmers than I are able to smooth that out. I think translating golog into pilog has been a great review of prolog and pilog and golog for me personally - hopefully useful for others on the list and elsewhere! I'm guessing picolisp/pilog for embedded systems will get more popular over time, as people see the advantages :-) Thanks for all the help! Doug --- On Mon, 7/4/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: pilog question To: picolisp@software-lab.de Date: Monday, July 4, 2011, 10:55 PM Hi Doug, Hmmm, what's the best answer for stackoverflow? (2 cons (- @F))) Tune the magic number to allow the correct unification bindings to reach up into rules which are calling this one, as far as needed, in a given application. The value of 2 here works for one test, but use 3 to make the other test work. Hmm, unsatisfactory ... :( I'm not very happy with the situation. It is a design flaw in Pilog. A cleaner solution would be to restructure the involved rules in such a way that the bindings are propagated up the normal way through the call hierarchy (and not from a deeply nested level directly to the top). I'm not sure how to do it in the case of golog, though. Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
pilog question
Consider these five pilog assertions. (be do ((Question @P) @S @S) (holds @P @S)) (be holds (@A @S) (restoreSitArg @A @S @F) (2 cons (- @F))) (be On (3 s0)) (be On (5 s0)) (be restoreSitArg ((On @N) @S (On @N @S))) I define some tests. (de t_1 () # ok (? holds On restoreSitArg (holds (On @N) s0)) ) (de t_2 () # ? (? do holds On restoreSitArg (do (Question (On @N)) s0 s0)) ) Run the first one. (t_1) 1 (holds (On @N) s0) 1 (restoreSitArg (On @N) s0 (On @N s0)) 1 (On 3 s0) @N=3 2 (On 5 s0) @N=5 - NIL Ok, in that first test, I asked for solutions to (holds (On @N) s0)), tracing on. Those are solutions I would expect, and they are correct. Second test. (t_2) 1 (do (Question (On @N)) s0 s0) 1 (holds (On @N) s0) 1 (restoreSitArg (On @N) s0 (On @N s0)) 1 (On 3 s0) @N=NIL 2 (On 5 s0) @N=NIL - NIL Now here, (holds (On @N) s0) is also attempted to be proved - because it is a sub-goal of (do (Question (On @N)) s0 s0). My question is, why the difference when (holds (On @N) s0) is proved? In one case solutions are found, in the other case it seems to miss the same solutions to the same goal. Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog question
Thanks Alex! I ended up taking it up to (5 cons (- @F)) to get the first simple elevator example working. Still a bit wonky in that respect. (Might have to adjust that magic unification level number per application!) But this definitely shows golog pilog is possible. Here's a page where golog.l and an example program (simple elevator) may be downloaded http://victoriafamilyrobotics.net/vfr1/golog Cheers, Doug --- On Mon, 7/4/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: pilog question To: picolisp@software-lab.de Date: Monday, July 4, 2011, 2:18 AM Hi Doug, (be do ((Question @P) @S @S) (holds @P @S)) (be holds (@A @S) (restoreSitArg @A @S @F) (2 cons (- @F))) Maybe this mechanism of explicitly specifying a number for the outer environment for unifying a clause is an an unfortunate limitation of Pilog. I don't know a better way, though. As I said, '2' tells the interpreter to unify in level 2, meaning the next level outside of 'holds'. So the query (? (holds (On @N) s0)) works, because you are unifying with the top level. On the other hand, if you'd defined (be holds (@A @S) (restoreSitArg @A @S @F) (3 cons (- @F)) ) i.e. specify level '3', the above query would not bind '@N', but the second query (? (do (Question (On @N)) s0 s0)) would bind it in turn. So as long as 'do' doesn't have '@N' in its environment (i.e. the level above 'hold'), specifying '2' in 'hold' causes the binding to get lost. I'm not sure if the Pilog interpreter could be easily rewritten to behave in a better way. Would it help to use '3' in 'hold'? Or, if you have cases where you also need '2', to use two versions of 'hold'? Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog question
Hmmm, what's the best answer for stackoverflow? (2 cons (- @F))) Tune the magic number to allow the correct unification bindings to reach up into rules which are calling this one, as far as needed, in a given application. The value of 2 here works for one test, but use 3 to make the other test work. ? --- On Mon, 7/4/11, Jakob Eriksson ja...@vmlinux.org wrote: From: Jakob Eriksson ja...@vmlinux.org Subject: Re: pilog question To: picolisp@software-lab.de Date: Monday, July 4, 2011, 5:04 AM http://stackoverflow.com/questions/6571200/pilog-assertion On Mon, Jul 04, 2011 at 12:26:46AM -0700, Doug Snead wrote: Consider these five pilog assertions. (be do ((Question @P) @S @S) (holds @P @S)) (be holds (@A @S) (restoreSitArg @A @S @F) (2 cons (- @F))) (be On (3 s0)) (be On (5 s0)) (be restoreSitArg ((On @N) @S (On @N @S))) I define some tests. (de t_1 () # ok (? holds On restoreSitArg (holds (On @N) s0)) ) (de t_2 () # ? (? do holds On restoreSitArg (do (Question (On @N)) s0 s0)) ) Run the first one. (t_1) 1 (holds (On @N) s0) 1 (restoreSitArg (On @N) s0 (On @N s0)) 1 (On 3 s0) @N=3 2 (On 5 s0) @N=5 - NIL Ok, in that first test, I asked for solutions to (holds (On @N) s0)), tracing on. Those are solutions I would expect, and they are correct. Second test. (t_2) 1 (do (Question (On @N)) s0 s0) 1 (holds (On @N) s0) 1 (restoreSitArg (On @N) s0 (On @N s0)) 1 (On 3 s0) @N=NIL 2 (On 5 s0) @N=NIL - NIL Now here, (holds (On @N) s0) is also attempted to be proved - because it is a sub-goal of (do (Question (On @N)) s0 s0). My question is, why the difference when (holds (On @N) s0) is proved? In one case solutions are found, in the other case it seems to miss the same solutions to the same goal. Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog: unification in variables that are clauses
Thanks Alex!!! We have golog in pilog. Er, I think :-) If not we're pretty dang close. Still trying to get the simple elevator example going from Knowledge in Action as a test. http://www.cs.toronto.edu/cogrobo/kia/simpleElevator http://books.google.com/books?q=%22proc%28goFloor%22 etc. After golog, maybe indigolog ? With miniPicoLisp at 100 to 150k (depending on platform, build options, etc.) and with pilog + golog loaded, it has to be the smallest footprint golog (with a respectable prolog and lisp under the hood), around. Might be right for embedded systems, I'm thinking. Cheers, Doug --- On Thu, 6/30/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: pilog: unification in variables that are clauses To: picolisp@software-lab.de Date: Thursday, June 30, 2011, 11:55 PM Hi Doug, the following works: (be a (3)) (be foo @C (2 - @C) ) : (? (foo (a @Z))) @Z=3 - NIL I spied what 'call/1' is doing ;-) That number magic with '2' is a bit tricky, I'm never sure what level to pick. Basically it tells the interpreter with which level it should unify (here, the level outside of 'foo'). Actually, you could also use (be foo (@C) (2 cons (- @C)) ) but this needs additional consing for the expression to match, so I would avoid it. It depends on what else is done with '@C'. Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
pilog: unification in variables that are clauses
--- On Wed, 6/29/11, Alexander Burger a...@software-lab.de wrote: So this should be written as (be holds (@A @S) (or ((restoreSitArg @A @S @F) (@ solve (list (- @F ((not (restoreSitArg @A @S @F)) (isAtom @A) (@ solve (list (- @A ) ) holds(A,S) :- restoreSitArg(A,S,F), F ; \+ restoreSitArg(A,S,F), isAtom(A), A. Now I think that the variables-as-clauses are getting me. In this rule, F might get bound to (unified with?) a clause with variables. For example, F might be bound to something like on(N,s0) When the F clause (now bound to on(N,s0)) is then proved, the N variable in on(N,s0) gets bound to something (3, for example). [trace] ?- holds(on(N), s0 ). Call: (6) holds(on(_G386), s0) ? creep Call: (7) restoreSitArg(on(_G386), s0, _G462) ? creep Exit: (7) restoreSitArg(on(_G386), s0, on(_G386, s0)) ? creep Call: (7) on(_G386, s0) ? creep Exit: (7) on(3, s0) ? creep Exit: (6) holds(on(3), s0) ? creep N = 3 . When I try to do this as (@ (solve (list (- @F in pilog, taking it through lisp , it prevents the values of the unified vars in @F from being used in the surrounding pilog rule. ? (? (holds (On @N) s0)) - T ? (solve '((holds (On @N) s0))) - (T) I wonder what I should do in this case, where a variable can be bound to a clause like that and the variables in that clause need to be bound, too? Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog: unification in variables that are clauses
Hi Alex, Thanks for your help! holds(A,S) :- restoreSitArg(A,S,F), F ; \+ restoreSitArg(A,S,F), isAtom(A), A. What I really want is this: (be holds (@A @S) (or ((restoreSitArg @A @S @F) @F) ((not (restoreSitArg @A @S @F)) (isAtom @A) @A) ) ) But the above (with a @F or @A for a clause) doesn't work. :-( This - (be holds (@A @S) (or ((restoreSitArg @A @S @F) (@ - @F)) ((not (restoreSitArg @A @S @F)) (isAtom @A) (@ - @A)) ) ) gives me the same results as using the (@ solve (list (- @F))) clause. Neither that nor (@ - @F) seem to unify variables in @F (which is bound to a clause) : (solve '((holds (On @N) s0))) - (T) Solves, but it also should tell me @N is bound to 3 Cheers, Doug --- On Thu, 6/30/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: pilog: unification in variables that are clauses To: picolisp@software-lab.de Date: Thursday, June 30, 2011, 12:27 AM On Thu, Jun 30, 2011 at 09:20:43AM +0200, Alexander Burger wrote: At least, a simple (@ . @A) should work. Doesn't it? Wait ... probably not a good idea. But this should work: (@ - @A) -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog: unification in variables that are clauses
Here's the issue boiled down to a simpler test case :-) First, swi-prolog: # cat t.pl a(3). foo(N) :- N. bar(a(X)) :- a(X). # swipl -f t.pl % /root/prolog/t.pl compiled 0.00 sec, 2,800 bytes ... ?- trace. true. [trace] ?- bar(a(Z)). Call: (6) bar(a(_G386)) ? creep Call: (7) a(_G386) ? creep Exit: (7) a(3) ? creep Exit: (6) bar(a(3)) ? creep Z = 3. [trace] ?- foo(a(Z)). Call: (6) foo(a(_G386)) ? creep Call: (7) a(_G386) ? creep Exit: (7) a(3) ? creep Exit: (6) foo(a(3)) ? creep Z = 3. In both cases, Z is correctly bound to 3. (The foo rule is the one we are interested in.) $ cat t.l (be a (3)) (be foo (@C) (@C - @C)) (be bar ((a @N)) (a @N)) ~/lisp/miniPicoLisp $ ./pil t.l : (? a bar (bar (a @X))) 1 (bar (a @N)) 1 (a 3) @X=3 - NIL That is correct: @X=3 ... however : : (? a foo (foo (a @X))) 1 (foo (a @X)) @X=NIL - NIL : But they should be the same, as @X = 3 is a (is the) solution. Moreover, in foo (a @X) is never even tried to be proven. So I think this test case demonstrates the issue. I still don't understand why this test of @A and/or @F for being non-NIL is necessary at all. This looks unusual to me. holds(A,S) :- restoreSitArg(A,S,F), F ; \+ restoreSitArg(A,S,F), isAtom(A), A. Me too. At first, I also thought they simply were testing for non-NIL there. But the intent of the 'naked' F and A clauses there seems to be this: If the data bound to F (or A) is itself a clause, then that clause gets proved and variables in it get unified/bound. So after the first restoreSitArg(A,S,F), F starts out as partially bound to a clause like on(N,s0) - which itself has an unbound variable, N. After F is proven, N is bound to 3 (because on(3,s0) is asserted elsewhere, and so N = 3 is a solution). A bit tricky, with variables bound to clauses which, in turn, have in them other unbound variables that may get bound. Using call/1 there (not sure if this is the way I should use it) gives similar results, where (a @X) is not attempted to be proven. $ cat t.l (be a (3)) (be foo (@C) # (@C - @C)) (call @C)) ~/lisp/miniPicoLisp $ ./pil t.l : (? a foo call (foo (a @X))) 1 (foo (a @X)) 1 (call (a @X)) - NIL : Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: Picked from HN: `First-class environments.'
Nice! I like that site rosettacode.org lots. (I stopped counting how many computer 'languages' I have assimilated over the years!) This example again shows off picolisp/pilog nicely. http://rosettacode.org/wiki/24_game/Solve#PicoLisp I think the perl solution may be a tad smaller. But most of us can't scan the perl one easily - the perl example there is dense and appears to be obfuscated. The picolisp/pilog example is not obfuscated and is smooth and clean IMO. :-) --- On Thu, 6/30/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: Picked from HN: `First-class environments.' To: picolisp@software-lab.de Date: Thursday, June 30, 2011, 5:50 AM Hi dexen + all, I accepted the challenge, and posted a new task to RosettaCode: http://rosettacode.org/wiki/First_class_environments The PicoLisp solution is quite clean, IMHO. Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: pilog: unification in variables that are clauses
The solve way seems to some closest to doing it. ~/lisp/miniPicoLisp $ cat t.l (be a (3)) (be foo (@C) # (@C - @C) # (call @C) (@ print (solve (list (- @C ) $ ./pil t.l : (? a foo (foo (a @Z))) 1 (foo (a @Z)) (((@Z . 3)))- T So I can see that it is correctly solving for @Z this way. But it doesn't take that binding (@Z = 3) back into the foo rule. -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
pilog question
I have a question about how or/2 works in pilog. Here's my test of or/2: ~/lisp/miniPicoLisp $ ./pil : (be a (3)) - a : (be b (3)) - b : (? (a 3)) - T : (be foo (@N) (or (a @N) (b @N))) - foo : (? foo a b or (foo 3)) 1 (foo 3) 1 (or (a 3) (b 3)) - NIL : Shouldn't that be T ? It seems to not to try to prove either (a 3) or (b 3). Am I using or/2 incorrectly in pilog? Here's that in swi-prolog for comparison: ~/prolog # cat t.pl a(3). b(3). foo(N) :- a(N) ; b(N). ~/prolog # swipl -f t.pl % /root/prolog/t.pl compiled 0.00 sec, 2,804 bytes Welcome to SWI-Prolog (Multi-threaded, 32 bits, Version 5.10.4) Copyright (c) 1990-2011 University of Amsterdam, VU Amsterdam SWI-Prolog comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under certain conditions. Please visit http://www.swi-prolog.org for details. For help, use ?- help(Topic). or ?- apropos(Word). ?- a(3). true. ?- foo(3). true . ?- -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
and/2 another pilog question
More pilog adventures! Alex, would it be possible to give pilog an and/2 rule similar to the or/2 rule that pilog has now? Sometimes refactoring pilog around the lack of an and/2 is a pain. Here's an example, where a rule has a kind of a logical expression (with prolog backup): holds(A,S) :- restoreSitArg(A,S,F), F ; \+ restoreSitArg(A,S,F), isAtom(A), A. I'd like to do something like: (be holds (@A @S) (or (and (restoreSitArg @A @S @F) (@ solve (list (- @F (and (not (restoreSitArg @A @S @F)) (isAtom @A) (@ solve (list (- @A))) ) ) ) Or's ok - but no and. (In pilog.) I think it is fully (?) backtracking over that expression in prolog too. But using a LISP and clause in a rule, like this: (@ and (solve ...) (solve ...)) Won't do backtracking over the legs of that (lisp) and. My solution so far is to make the parts of the conjunction be separate rules to get the desired backtracking: (be holds (@A @S) (or ((holdsRestoreSitArg @A @S @F) (holdsNotRestoreSitArg @A @S @F (be holdsRestoreSitArg (@A @S @F) (restoreSitArg @A @S @F) (@ solve (list (- @F))) ) (be holdsNotRestoreSitArg (@A @S @F) (not (restoreSitArg @A @S @F)) (isAtom @A) (@ solve (list (- @A))) ) (The (@ solve (list (- @x))) -type clause here is trying to treat a naked variable as a rule clause - a workaround because you can't have a pilog variable in place of a rule clause - or at least it doesn't get instantiated like prolog if you try, rather. But this is a separate issue from the and/2 request.) Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: and/2 another pilog question
Ha! The and is implicit ... I should seen that - the or/2 example there was a clue. Thanks for the assistance!! I feel close to getting golog (and elevator example) working. Cheers, Doug --- On Wed, 6/29/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: and/2 another pilog question To: picolisp@software-lab.de Date: Wednesday, June 29, 2011, 9:12 PM Hi Henrik, What is select ( http://software-lab.de/doc/refS.html#select/3 ) using internally for the filter part? Seems to me the filter part is a kind of implicit and. Right. As I wrote in the previous mails, all clauses in a Prolog rule (not only in 'select/3') are combined by an implict AND. Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: golog in minipicolisp
--- On Sun, 6/26/11, Doug Snead semaphore_2...@yahoo.com wrote: (be isAtom (@A) (not (equal @A (Neg @W))) (or (equal @A (And @W1 @W2))) (or (equal @A (If @W1 @W2))) (or (equal @A (Is @W1 @W2))) (or (equal @A (Or @W1 @W2))) (or (equal @A (some @X @W))) (or (equal @A (all @X @W I mean: (be isAtom (@A) (not (or (equal @A (Neg @W)) (equal @A (And @W1 @W2)) (equal @A (If @W1 @W2)) (equal @A (Is @W1 @W2)) (equal @A (Or @W1 @W2)) (equal @A (some @X @W)) (equal @A (all @X @W) :-) -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: tracing pilog
Thanks Alex!The more I look at it, the more I like pilog for golog because golog uses prolog-ish backtracking so heavily. When I look at this attempt to make a lua golog for example http://drops.dagstuhl.de/opus/volltexte/2010/2631/pdf/10081.Ferrein.2631.pdf , I think a picolisp/pilog -based golog might be a better approach than using lua for this, at least. (And by translating the swi prolog golog into pilog, I seem to be able to understand golog programs better, at least.) Cheers, Doug --- On Sun, 6/26/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: tracing pilog To: picolisp@software-lab.de Date: Sunday, June 26, 2011, 12:31 AM Hi Doug, What's the best way to trace pilog? Pilog clauses can indeed be traced. Unfortunately, I don't find a good description at the moment. The reference of '?' (and of 'prove' which is the internal machinery of the query front end '?') just briefly mentions it: http://software-lab.de/doc/ref_.html#? http://software-lab.de/doc/refP.html#prove It works by simply passing the names of the clauses you want to trace right after the '?' (i.e. the ['sym' ..] arguments). A simple example, using the 'append' clause: Without tracing : (? (append (a b c) @X (a b c d e f))) @X=(d e f) - NIL With tracing : (? append (append (a b c) @X (a b c d e f))) 2 (append (a b c) @Y (a b c d e f)) 2 (append (b c) @Y (b c d e f)) 2 (append (c) @Y (c d e f)) 1 (append NIL (d e f) (d e f)) @X=(d e f) The trace output lists all clauses that _matched_, and always starts with a number indicating _which_ of the clauses in the definition matched. The definition of 'append' is : (rules 'append) 1 (be append (NIL @X @X)) 2 (be append ((@A . @X) @Y (@A . @Z)) (append @X @Y @Z)) - append The symbolic arguments to '?' can of course also specify clauses somewhere deeply nested in the Pilog definitions, not just those passed to '?' on the top level. Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: tracing pilog
This is exactly what I was looking for, thanks! --- On Sun, 6/26/11, Alexander Burger a...@software-lab.de wrote: It works by simply passing the names of the clauses you want to trace right after the '?' (i.e the ['sym' ..] arguments). ... With tracing : (? append (append (a b c) @X (a b c d e f))) -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: embedding minipicolisp
Hi Alexander, Ha! That last code snippet got me to where I needed to get :-) THANKS!! (Now when I close my eyes I see pilog - what does that mean?) Cheers, Doug --- On Wed, 6/22/11, Alexander Burger a...@software-lab.de wrote: From: Alexander Burger a...@software-lab.de Subject: Re: embedding minipicolisp To: picolisp@software-lab.de Date: Wednesday, June 22, 2011, 12:28 AM Hi Doug, I can then doPrint() on what doRead() returns - although I don't see the customary parentheses around the list I hand that to doEval(). Calling the high-level Lisp function 'doEval' is not a good idea, because it first evaluates its arguments, which is not what is needed here. The same is the case for 'doRead', but it doesn't do much harm if you just pass it NIL (i.e. no arguments). And also for 'doPrint': If you pass it something other than a number, if will print not that item but its value. Better just call 'print()' for debugging. So instead of calling 'doEval', you could do any res; cell c1; Push(c1, doRead(Nil)); x = EVAL(data(c1)); drop(c1); return x; doEval() returns a value (an any pointer, which when cast to a long is 6 which seems odd). '6' is a valid 'any', the number '1': num xx10 2 0110 Cheers, - Alex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: embedding minipicolisp
Ok, here's my first attempt at using minipicolisp as a library. (Leaving aside
the issue of get/put for now. Basically replacing main() with:
#if !LIBRARY
/*** Main ***/
int main(int ac, char *av[]) {
char *p;
. . .
#else
// we're building as a library
void miniPicoLisp_init( const char * home_dir )
{
heapAlloc();
initSymbols();
if (home_dir) {
int l = strlen(home_dir) + 1;
Home = malloc( l );
memcpy(Home, home_dir, l );
Home[ l ] = '\0';
}
Reloc = Nil;
InFile = stdin, Env.get = getStdin;
OutFile = stdout, Env.put = putStdout;
// ??
ApplyArgs = cons(cons(consSym(Nil,0), Nil), Nil);
ApplyBody = cons(Nil,Nil);
}
any miniPicoLisp_dostring( char *str )
{
return load(NULL, 0, mkStr(str));
}
#endif
And then testing it with this:
#include pico.h
extern void miniPicoLisp_init( const char * home_dir );
extern any miniPicoLisp_dostring( char *str );
#define MAX_STR 1024
int main(int ac, char *av[]) {
char buf[MAX_STR];
miniPicoLisp_init( NULL );
while (fgets(buf, MAX_STR, stdin) != NULL) { // read
any v = miniPicoLisp_dostring( buf ); // eval
doPrint( v ); // print
}
return 0;
}
Doug
--- On Sat, 6/18/11, Alexander Burger a...@software-lab.de wrote:
From: Alexander Burger a...@software-lab.de
Subject: Re: embedding minipicolisp
To: picolisp@software-lab.de
Date: Saturday, June 18, 2011, 11:21 PM
Hi Doug,
I'd like to embed minipicolisp in C applications I'd
like to be able
to create a minipicolisp environment, and then pass
strings to it
(commands) for execution. It would also be great to be
able get back the
results in a string.
The mechanism in PicoLisp to execute a string is 'load'. If
an argument
to 'load' starts with a hyphen, the rest is interpreted as
an expression
(without the surrounding parentheses). This mechanism is
normally used
to run expressions from the command line.
The lower level functions for that are 'str', both for
converting a
transient symbol to a list (without the surrounding
parentheses) and to
do the reverse, and EVAL() (which is actually not a
function but a
macro).
Unfortunately, there is no way currently to convert a C
string directly
to an executable Lisp expression (without the intermediate
transient
symbol).
In lua, you'd do something like,
L = lua_open();
Here you'd need a function that initialized the PicoLisp
VM, as is done
by main(), i.e. calls heapAlloc(), initSymbol() etc.
luaL_dostring(L, str_in );
str_out = l_results_to_string(L);
You need a C function similar to 'loadAll()'. It would call
'mkStr()' on
the C string to create the transient symbol, then load() to
run the
expression.
Better would be to cook something more specialized, similar
to parse(),
to avoid the intermediate transient symbol step.
To pack the result into a C string, you need a mechanism
similar to
doStr().
You define a function which writes a single byte into a C
character
array,
void putCharC(int c) {
*CBytePtr++ = c;
}
and store a pointer to that function into 'Env.put':
void (*putSave)(int);
putSave = Env.put;
Env.put = putCharC;
print(result);
Env.put = putSave;
Hope this helps.
Cheers,
- Alex
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
golog in minipicolisp ?
Alexander (or anyone!), I'm looking at something called golog http://www.cs.toronto.edu/cogrobo/main/systems/index.html it is a A high-level agent programming language... based on Prolog (usually Eclipse, SWI, LPA, and Quintus.) An interpreter in SWI Prolog can be found here. http://www.cs.toronto.edu/cogrobo/Systems/golog_swi.pl An interpreter in ECLIPSE Prolog can be found here. http://www.cs.toronto.edu/cogrobo/Systems/gologinterpreter.pl The .pl prolog gologs above look (deceptively) simple. I don't know pilog well enough to be able to tell if there is something about gologinterpreter.pl (or golog_swi.pl, same thing) that I will NOT be able to do in minipicolisp's pilog. I'd like to try to converting gologinterpreter.pl into pilog (especially for minipicolisp). I am hoping that this will be both possible and straightforward. Any thoughts? Cheers, Doug -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: script for generating tags for picolisp
Here's one shell script I used to use for that (I called it lisptags) ...
hopefully it still works :-)
#!/bin/sh
# make a tags file for pico lisp source files.
# use:
# lisptags foo.l bar.l baz.l ... bof.l
# generate the file 'tags'
# [based on lisptags csh script by John Foderaro, c.1982]
awk '
/^[ \t]*\([bd][em][ \t]/ {
n=$2;
gsub(/^\(/,,n);
gsub(/\)$/,,n);
print n \t FILENAME \t?^ $0 $?;
}
/^[ \t]*\(class[ \t]/ {
n=$2;
gsub(/^\(/,,n);
gsub(/\)$/,,n);
print n \t FILENAME \t?^ $0 $?;
if (n ~ /^\+/) {
gsub(/^\+/,,n);
print n \t FILENAME \t?^ $0 $?;
}
}
' $* | sort tags
= = =
--- On Tue, 8/17/10, Edwin Eyan Moragas e...@yndy.org wrote:
From: Edwin Eyan Moragas e...@yndy.org
Subject: script for generating tags for picolisp
To: picolisp@software-lab.de
Date: Tuesday, August 17, 2010, 12:05 AM
Hi all,
anybody who made an attempt to create such kind of script?
or maybe a
resource somewhere in the interwebs for this?
cheers,
/e
--
UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe
--
UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe
Re: Concurrency
--- On Fri, 10/9/09, Tomas Hlavaty t...@logand.com wrote:=0A How many pro= cesses do you need?=A0 I can fork about 500=0A from one parent=0A (limite= d by my OS set up).=A0 And because picolisp is,=0A well, pico, it=0A do= esn't take much memory (picolisp uses less that 2MB).=0A=0AAnd if picolisp = is built as a shared object/dll, then additional picolisp processes may con= sume even less memory. =0A=0ADoug=0A=0A=0A=0A -- UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe
