Re: building on macos
Regarding the dependency on llvm-16, I get that during the build process with llvm-10: (cd src; make) /opt/local/libexec/llvm-10/bin/llc: error: /opt/local/libexec/llvm-10/bin/llc: picolisp.bc: error: Invalid value (Producer: 'LLVM16.0.6' Reader: 'LLVM 10.0.1') make: *** [../bin/picolisp] Error 1 Jean-Christophe > On Jan 12, 2025, at 11:42, Jean-Christophe Helary > wrote: > > Hello everyone! > > Some updates that need confirmation. > > https://picolisp.com/wiki/?alternativeMacOSRepository > > It looks like picolisp now requires llvm-16. > > Also, for a reason that escapes me (maybe a bug on the part of the MacPort > team, but I need to investigate), port installs llvm-16 binaries with the > -mp-16 suffix *but* does not create aliases to the non-suffixed version. > > That generates a number of “command not found” errors, namely: > - llvm-config > - opt > - llvm-link > - llc > > I’ll check that with the MacPort teams before proposing a fix. But it’s weird > because the non-suffixed aliases seemed to have been produced in previous > versions of the llvm port. > > et voilà ! > > Jean-Christophe > >> On Aug 14, 2023, at 22:36, l@tlo wrote: >> >> >> >>> On Aug 14, 2023, at 15:15, l@tlo wrote: >>> >>> One year later... >>> >>> (I don't even remember sending that older mail... 😅) >>> >>> >>> Building on macOS (13) seems to work fine with the following instructions: >>> >>> https://picolisp.com/wiki/?alternativeMacOSRepository >> >> I just noticed the reply by Louis Abraham: >> >> https://picolisp.com/wiki/-A725.html >> >> For readability it might be better to merge the two documents into one: the >> brew style and the macport style. >> >> Louis, you seem to mean that your version works with M1 machines. Have you >> tried it on Intel machines? >> >> JC >> >>> >>> *and* >>> >>> Mike's Makefile here: >>> >>> https://git.envs.net/mpech/pil21-tests/raw/branch/master/Makefile.macos >>> >>> >>> One question to Mike: >>> >>> Would it be possible that you merge your makefile with Alex's most recent >>> and add conditionals to check whether the system is macOS? >>> >>> That way, we would not need any special instructions to build on macOS :) >>> >>> >>> If not, I'll update the instructions and will add a link to your Makefile. >>> >>> >>> And then, Mia's blog article could be updated to reflect the current status >>> :) >>> >>> https://picolisp-explored.com/how-to-install-picolisp >>> >>> >>> Jean-Christophe >>> On May 30, 2022, at 13:39, Jean-Christophe Helary wrote: > On May 30, 2022, at 18:18, Alexander Burger wrote: > > Hi Jean-Christophe, > >> It looks like I'm slowly getting somewhere... >> ... > > Great! :) > > >> ./pil: line 2: exec: >> /Users/suzume/Documents/Repositories/pil21+/src/../bin/picolisp: cannot >> execute: No such file or directory > > That's an easy one. It tries to bootstrap, but bin/picolisp does not > exist yet. > For that reason the pre-build *.ll files are included in the distro. Just > (re)install these. Yes, but now, I'm back to my readline errors... ➜ pil21+ git:(master) ✗ (cd src; make) clang -O3 -w -c -o lib.bc -D_OS='"Darwin"' -D_CPU='"x86_64"' `pkg-config --cflags libffi` -emit-llvm lib.c lib.c:119:4: error: use of undeclared identifier 'rl_catch_signals' rl_catch_signals = 0; ^ lib.c:121:4: error: use of undeclared identifier 'rl_input_available_hook' rl_input_available_hook = rlAvail; ^ (etc.) I'm using this in .profile: ## libffi export PKG_CONFIG_PATH="/usr/local/opt/libffi/lib/pkgconfig:$PKG_CONFIG_PATH" export LDFLAGS="-L/usr/local/opt/libffi/lib:$LDFLAGS" export CPPFLAGS="-I/usr/local/opt/libffi/include:$CPPFLAGS" ## openssl export PATH="/usr/local/opt/openssl@3/bin:$PATH" export LDFLAGS="-L/usr/local/opt/openssl@3/lib:$LDFLAGS" export CPPFLAGS="-I/usr/local/opt/openssl@3/include:$CPPFLAGS" ## readline export PKG_CONFIG_PATH="/usr/local/opt/readline/lib/pkgconfig:$PKG_CONFIG_PATH" export LDFLAGS="-L/usr/local/opt/readline/lib:$LDFLAGS" export CPPFLAGS="-I/usr/local/opt/readline/include:$CPPFLAGS" ## llvm export PATH="/usr/local/opt/llvm@11/bin:$PATH" export LDFLAGS="-L/usr/local/opt/llvm@11/lib:$LDFLAGS" export CPPFLAGS="-I/usr/local/opt/llvm@11/include:$CPPFLAGS" JC :( > > > -- > Jean-Christophe Helary > https://sr.ht/~brandelune/ > @jchelary@sciences.social > -- Jean-Christophe Helary https://sr.ht/~brandelune/ @jchelary@sciences.social -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
How to use enum? to check for existence in sparse array
Hi Alex, Is this a bug or is my understanding of how to use enum? from the documentation not correct. I have a little project where I would like to use 'enum. I was using 'idx before, but 'enum looks like a much better fit. My current problem is that in a sparse array I am getting false positives as to whether a node exists when checking with enum? How do I use 'enum? to check for entry existence? Using the (val (enum ..)) and checking for NIL also modifies the tree does it not. It is also much slower than just 'enum? Regards, /Lindsay # () # Generate a 'Hash' value (based on Lehmer pseudo-random algo (de Hasher (Val Seed) (default Seed 7337) (let (State Seed Lst (if (lst? Val) Val (list Val))) (for C Lst (setq State (% (* (* C State) 279470273) 4294967291)) ) ) ) # Generate some hash values (de TestEnum (N) (off *HT) (for I N (let (Val (Hasher I)) (if (enum? '*HT Val) (prinl "! Duplicate: " I "; " Val "; " (enum? '*HT Val)) (set (enum '*HT Val) I) (prinl "+ Added: " I "; " Val "; " (enum? '*HT Val)) ) ) ) ) # () # Why does enum? report 'duplicate' hash value : (TestEnum (** 2 3)) + Added: 1; 1773995194; ! Duplicate: 2; 3547990388; 1 + Added: 3; 1027018291; + Added: 4; 2801013485; + Added: 5; 280041388; ! Duplicate: 6; 2054036582; 3 + Added: 7; 3828031776; + Added: 8; 1307059679; -> NIL # The hash values are all unique : (for C (** 2 3) (printsp C (Hasher C)) (prinl) ) 1 1773995194 2 3547990388 3 1027018291 4 2801013485 5 280041388 6 2054036582 7 3828031776 8 1307059679 -> NIL # 'enum shows TestEnum didn't add the value, but also creates it doesn't it? : (for C (** 2 3) (println (val (enum '*HT (Hasher C) 1 NIL 3 4 5 NIL 7 8 -> 8 P.S. For the integers between 1 and (2 ** 20) (the extent of my testing so far) the Hasher functions generates unique values that are nicely distributed and work well in an 'idx tree. As 'enum seems much faster (and simpler) than 'idx, I would like to use that, with an 'assoc list in the value for any potential collisions. Note: the hasher currently handles lists of numbers as well.
Re: How to use enum? to check for existence in sparse array
> Is this a bug or is my understanding of how to use enum? from the > documentation not correct. > > My mistake! The correct syntax is (enum? *HT Val) and not (enum? '*HT Val). I was quoting the tree name. Doh! Regards, Lindsay Note: I am still not entirely clear why 'enum? can return unexpected (at least to me) results in the sparse array case (see below) Is it because 'enum precreates key nodes in the tree for a sparse array in order to keep it balanced?. e.g. : (pp 'TestEnum) (de TestEnum (N) (off *HT) (for I N (let (Val (Hasher I)) (if (enum? *HT Val) (prinl "! Duplicate: " I "; " Val "; " (sym (enum? '*HT Val)) "; " (sym (val (enum '*HT Val))) ) (set (enum '*HT Val) I) ) ) ) ) -> TestEnum : (TestEnum (** 2 16)) ! Duplicate: 21487; 25853; NIL; NIL ! Duplicate: 42974; 51706; (NIL (NIL NIL NIL (NIL NIL NIL (NIL (NIL (NIL (NIL (NIL (NIL NIL NIL (NIL NIL NIL NIL NIL NIL NIL NIL NIL NIL 11011 NIL NIL NIL (NIL (NIL NIL NIL NIL NIL NIL NIL (NIL (NIL NIL NIL NIL NIL NIL NIL NIL NIL (NIL NIL 28854)) NIL (NIL NIL NIL NIL NIL NIL NIL NIL NIL (NIL (NIL (NIL NIL NIL (NIL NIL NIL NIL NIL NIL 3644)); NIL ! Duplicate: 50121; 260192; NIL; NIL ! Duplicate: 64461; 77559; NIL; NIL -> 65536 > >
Re: How to use enum? to check for existence in sparse array
> > Note: I am still not entirely clear why 'enum? can return unexpected (at > least to me) results in the sparse array case (see below) > Is it because 'enum precreates key nodes in the tree for a sparse array in > order to keep it balanced?. > > Ok. I think I get the structure. (val (enum? *HT Val) is the test I need to use and is very fast! My apologies for the email noise as I worked through this. /Lindsay (de TestEnum (N) (off *HT) (for I N (let (Val (Hasher I)) (if (val (enum? *HT Val)) (prinl "! Duplicate: " I "; " Val "; " (sym (enum? '*HT Val)) "; " (sym (val (enum '*HT Val))) ) (set (enum '*HT Val) I) ) ) ) ) : (TestEnum (** 2 16)) -> 65536 : (TestEnum (** 2 18)) -> 262144 : (TestEnum (** 2 20)) -> 1048576
(em)
Is it my (macOS) system or (em) does not work ? I guess it must be a library that is not loaded at launch ? pil -em + !? (em) em -- Undefined -- Jean-Christophe Helary https://sr.ht/~brandelune/ @jchelary@sciences.social -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Connecting stderr
Hi, When calling, or piping, to an external process from picolisp, is there a way to connect the stderr of that process on another channel as well, so that it can be read, and distinguished, from stdout. /Lindsay
Subscribe
Hello Emon Sahariar :-) You are now subscribed -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: How to set initial job environment values?
> > > > 'curry' is a bit overkill here. I would directly 'fill' a 'job': > Thanks! 'fill' is indeed the function I was looking for here. Working with 'nth', I did learn more about the picolisp machine and list internals. /Lindsay z
Re: Delayed Mails
On Tue, Jan 28, 2025 at 9:47 AM Alexander Burger wrote: > Sorry for the five delayed Mails! It was my fault :( > > ☺/ A!ex 👍
Re: How to set initial job environment values?
On Sun, Jan 26, 2025 at 12:51:10AM -0800, Lindsay Lawrence wrote: > Is there a simpler way to set the curry environment's State value to Seed > than what I am doing here? > > Regards, > Lindsay > > (de randomLehmer (Seed) >(let > (Fn > (curry > ((State . 1)) > NIL > (setq State (% (* State 279470273) 4294967291)) ) ) > (set > (nth Fn 2 2 2) > (cons > 'State > (if Seed Seed (inc (abs (rand ) ) > Fn ) ) > > > : (pretty (randomLehmer 97)) > (NIL >(job '((State . 97)) > (setq State (% (* State 279470273) 4294967291)) ) )-> ")" 'curry' is a bit overkill here. I would directly 'fill' a 'job': (de randomLehmer (Seed) (let @Seed (or Seed (inc (abs (rand (fill '(() (job '((State . @Seed)) (setq State (% (* State 279470273) 4294967291)) ) ) ) ) ) : (pretty (randomLehmer 97)) (NIL (job '((State . 97)) (setq State (% (* State 279470273) 4294967291)) ) )-> ")" ☺/ A!ex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
picolisp db encryption?
Question: Is it possible to encrypt picolisp dbs on the fly? For example, to hook into the low level functions that do block reads and writes in order to apply encryption? For example, it would be a useful option, when setting up a 'pool' to specify a key that is used to decrypt/encrypt block reads/writes to the underlying storage. Once set up, this action would be transparent to the application. /Lindsay
Re: picolisp db encryption?
On Tue, Jan 28, 2025 at 11:10:25AM -0800, Lindsay Lawrence wrote: > Question: Is it possible to encrypt picolisp dbs on the fly? For example, > to hook into the low level functions that do block reads and writes in > order to apply encryption? Not so easily. If it were only for reading from such a DB, it would be possible via the '*Ext' mechanism. With that, almost anything can be read as a database (for example PilBox is doing that to access the Java Objects in the Android runtime). But for writing I see no easy way at the moment. > For example, it would be a useful option, when setting up a 'pool' to > specify a key that is used to decrypt/encrypt block reads/writes to the > underlying storage. Once set up, this action would be transparent to the > application. Also, I'm not sure if this is really useful, as it could always be locally intercepted. Needs more thinking ... ☺/ A!ex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
unsubscribe
Good bye Alex Williams :-( You are now unsubscribed -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: Connecting stderr
On Sat, Jan 25, 2025 at 8:56 AM Lindsay Lawrence <
lawrence.lindsayj...@gmail.com> wrote:
> Is there a straightforward way from PicoLisp, when executing an external
> process, to get stderr on a separate channel?
> I'd like to do something like (pipe) where I can connect stdin and stdout,
> but have an additional channel for the child's stderr
>
> After a bit of hacking, I've come up with the code below:
It uses pipe, redirects stderr of the Exe to a file and returns a list with
the stdout lines in the car and stderr in the cadr
I'm pretty happy with it :)
/Lindsay
(de runExe (Cmds . Exe)
(use (LOut LErr)
(let (ErrFile (tmp "exerr." (rand)))
(finally (call "rm" ErrFile)
(pipe
(err ErrFile (out Exe (mapc prinl Cmds)))
(setq LOut (make (while (line T) (link @ )
(in ErrFile
(setq LErr (make (while (line T) (link @ ) ) )
(list LOut LErr) ) )
which works something like this (note the error in the last command sent)
: (mapc '((X) (mapc println X))
(runExe '(".mode quote"
"select strftime('%s', 'now'), date('now');"
"select date();"
"select time();"
"select datetim();"
".exit")
sqlite3 "file:mdb?mode=memory" ))
"'1737836168','2025-01-25'"
"'2025-01-25'"
"'20:16:08'"
"Error: near line 5: no such function: datetim"
-> "Error: near line 5: no such function: datetim"
Re: How to use enum? to check for existence in sparse array
On Mon, Jan 13, 2025 at 10:58:55PM -0800, Lindsay Lawrence wrote: > Ok. I think I get the structure. > (val (enum? *HT Val) is the test I need to use and is very fast! Right. Sorry once more for the late reply! > My apologies for the email noise as I worked through this. No problem! ;) ☺/ A!ex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Updated wiki example
Hi,
https://picolisp.com/wiki/?RunExe
I optimized the code used in the example I previously added to the wiki. Being
able to pipe 1M inserts into a sqlite db in ~2.5secs from picolisp is all
kinds of awesome.
I also added a link to an additional example:
https://github.com/thinknlive/picolisp-lisp-basics/blob/master/sqlite.l
In the additional example, I have added a couple of functions to quote sql
parameters
and a (just) good enough parser for results returned as json, allowing
reliably correct handling of unicode and escaped characters on inserts and
queries. (see bench code below)
Happy Hacking
/Lindsay
: (bench (let (Sql '(
".mode quote"
"drop table if exists test;"
"create table test(id integer primary key, name text, cnt int default 1);"
"create unique index uix_test_name on test(name);"
))
(setq Sql
(append
Sql
'("begin transaction;")
(make (do 1 (link (pack
"insert into test (name) values (" (sqlQuote (makeRndName 4)) ") on
conflict (name) do update set cnt = cnt + 1;"
'("commit;")
'("select json_array(id, name, cnt, (hex(randomblob(8 from test
where cnt > 1;")
'("select json_array('CollisionCount', count(*)) from test where cnt
> 1;")
'("select json_array('RecordCount', count(*)) from test;")))
(bench (mapc '((R) (let (Row (head -2 (nth (chop R) 3))
Flds (fldParseQ Row))
(prin R " --> ") (print Flds) (prinl " : " (length
Flds
(car (runExe Sql sqlite3 "testdb.sqlite"))) ]
'[31,"99\\,",2,"0A0427635B128F1B"]' --> ("31" "99\\," "2"
"0A0427635B128F1B") : 4
'[32,"6,👍3",2,"0A15EE4D50BA7DED"]' --> ("32" "6,👍3" "2"
"0A15EE4D50BA7DED") : 4
'[142,"👍238",2,"5FB33CBC8DC3B366"]' --> ("142" "👍238" "2"
"5FB33CBC8DC3B366") : 4
'[266,"CIᘐU",2,"4AB7193F10CEBB55"]' --> ("266" "CIᘐU" "2"
"4AB7193F10CEBB55") : 4
'[472,"\np👍\"",2,"979F6057BCD7C97B"]' --> ("472" "^Jp👍\"" "2"
"979F6057BCD7C97B") : 4
'[990,"\\n👍3",2,"5F3F9E2031EA3C52"]' --> ("990" "\\n👍3" "2"
"5F3F9E2031EA3C52") : 4
'[1546,"\n1,6",2,"0CB9CE64B32AFE6D"]' --> ("1546" "^J1,6" "2"
"0CB9CE64B32AFE6D") : 4
'[1836,",\\5ᘐ",2,"9FD2D32906D5081A"]' --> ("1836" ",\\5ᘐ" "2"
"9FD2D32906D5081A") : 4
'[2188,"8FZ👍",2,"C3D2C1661145BE91"]' --> ("2188" "8FZ👍" "2"
"C3D2C1661145BE91") : 4
'[2239,"\"9ᘐ\"",2,"C61BF491F245F996"]' --> ("2239" "\"9ᘐ\"" "2"
"C61BF491F245F996") : 4
'[3106,",\n2L",2,"C33A015041A908B0"]' --> ("3106" ",^J2L" "2"
"C33A015041A908B0") : 4
'[3731,"\\👍\"9",2,"AD7232A882E4BD03"]' --> ("3731" "\\👍\"9" "2"
"AD7232A882E4BD03") : 4
'[3748,"\n\",\\",2,"700011DF8C20EEED"]' --> ("3748" "^J\",\\" "2"
"700011DF8C20EEED") : 4
'[5288,"3g👍\\",2,"D5F23C683C48CA23"]' --> ("5288" "3g👍\\" "2"
"D5F23C683C48CA23") : 4
'[6579,"\"\"\n,",2,"8C594685D3E210CF"]' --> ("6579" "\"\"^J," "2"
"8C594685D3E210CF") : 4
'["CollisionCount",15]' --> ("CollisionCount" "15") : 2
'["RecordCount",9985]' --> ("RecordCount" "9985") : 2
0.135 sec
0.158 sec
-> 2
:
Re: Connecting stderr
Hi, Re-asking the question: Is there a straightforward way from PicoLisp, when executing an external process, to get stderr on a separate channel? I'd like to do something like (pipe) where I can connect stdin and stdout, but have an additional channel for the child's stderr /Lindsay On Wed, Jan 15, 2025 at 10:02 AM Lindsay Lawrence < lawrence.lindsayj...@gmail.com> wrote: > Hi, > > When calling, or piping, to an external process from picolisp, is there a > way to connect the stderr of that process on another channel as well, so > that it can be read, and distinguished, from stdout. > > > /Lindsay >
How to set initial job environment values?
Hi, Is there a simpler way to set the curry environment's State value to Seed than what I am doing here? Regards, Lindsay (de randomLehmer (Seed) (let (Fn (curry ((State . 1)) NIL (setq State (% (* State 279470273) 4294967291)) ) ) (set (nth Fn 2 2 2) (cons 'State (if Seed Seed (inc (abs (rand ) ) Fn ) ) : (pretty (randomLehmer 97)) (NIL (job '((State . 97)) (setq State (% (* State 279470273) 4294967291)) ) )-> ")" : > >
Re: (em)
On Tue, Jan 14, 2025 at 09:44:10AM +, Jean-Christophe Helary wrote: > Is it my (macOS) system or (em) does not work ? > > I guess it must be a library that is not loaded at launch ? > > pil -em + > !? (em) > em -- Undefined AFAIK (em) was supported only on pil64, not on current pil21. Not sure what would be needed to port it. ☺/ A!ex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Delayed Mails
Sorry for the five delayed Mails! It was my fault :( ☺/ A!ex -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
Re: Subscribe
> On Jan 13, 2025, at 20:21, Alexander Burger wrote: > > On Sun, Jan 12, 2025 at 02:53:49AM +, Jean-Christophe Helary wrote: >> Hello Jean-Christophe Helary >> :-) >> You are now subscribed >> >> >> Am I subscribed ? > > Yes :) Though this answer is probably not helpful, because you either > don't see it, or you received your own posts already. :) > I have no info about the Mac issues. I don’t know about llvm-16 being required, it is just a message I got from the make process that hinted at that. As for the aliases, it looks like something is fishy, but I am not sure yet if it’s my system of the way the distribution does (not) inform users about the instructions required to create aliases. When this is sorted out, I’ll update the wiki. -- Jean-Christophe Helary https://sr.ht/~brandelune/ @jchelary@sciences.social -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
