»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
b2gills m: print $0 if "[abc]def" ~~ / \[ ( <-[\]]>+ ) \] / 01:03
camelia abc
timotimo m: print $0 if "[abc]def" ~~ / '[' ( <- ']' > ) ']' / 01:04
camelia 5===SORRY!5===
Unrecognized regex metacharacter < (must be quoted to match literally)
at <tmp>:1
------> 3print $0 if "[abc]def" ~~ / '[' ( <-7⏏5 ']' > ) ']' /
Unrecognized regex metacharacter - (must be quoted to match literally)
at …
timotimo m: print $0 if "[abc]def" ~~ / '[' ( <-']'> ) ']' /
camelia 5===SORRY!5===
Unrecognized regex metacharacter < (must be quoted to match literally)
at <tmp>:1
------> 3print $0 if "[abc]def" ~~ / '[' ( <-7⏏5']'> ) ']' /
Unrecognized regex metacharacter - (must be quoted to match literally)
at <t…
timotimo mhm
b2gills m: print $0 if "[abc]def]" ~~ / '[' ~ ']' ([ <!before ']'> . ]+) / 01:39
camelia abc
TimToady m: print $0 if "[abc]def]" ~~ / '[' ~ ']' (.*?) / 01:44
camelia abc
TimToady that seems simpler
and is not exactly backtracking, more like forward tracking... 01:45
timotimo ya 01:45
TimToady though probably doesn't work in a grammar that defines FAILGOAL 01:46
timotimo TimToady: do we have a clue how we can emit regexes that don't actually compile to regexes? 01:49
for example, on the mailing list we just had someone want s|.*'/'||
that'd totally work as just a substr with an rindex
TimToady not yet, but I'm working my way into regex/grammar optimization, now that I'm mostly done with braids 02:01
timotimo yup 02:10
kalkin- hi #perl6 02:22
How do i read from a file descriptor which I obtained from a native call?
timotimo well, you can definitely use nativecall for regular libc functions like fprintf 02:23
we do have a "give me the native descriptor for this perl6-level IO thingie" 02:24
not sure about the other side
well, you can "/proc/self/fd/$num".IO.open(...)
but only on linux
kalkin- timotimo: i think the other side is missing
timotimo: TIL about fd/$num 02:25
timotimo: is linux only anyway, thanks! 02:26
ups
timotimo i hope it actually works out! :) 02:28
jiang p6:say 3 02:39
avuserow p6: say 3 # need a space after the colon? 02:40
camelia 3
jiang p6: say 'oh, just like this ;-)'; 02:41
camelia oh, just like this ;-)
kalkin- /proc/self/fd/$num doesn't exist, but /proc/$*PID/fd/$num does (checked via shell during runtime), but when accessing it i get permission denied 02:42
timotimo well, accessing it from a different thread surely won't work 02:43
how were you checking for it?
wow, just .say for dir /proc/self/fd gives me 0 through 13 02:44
impressive
user_ timotimo: "/proc/$*PID/fd/$inotify.fd()".IO; 02:45
timotimo huh, 0, 1, 2, 9, 11, and 12 are all /dev/tty 02:45
and one /dev/null, too
a few precomps are mmapped, that's not so surprising that they'd have fds 02:46
user_ btw the fd in question is shown in zsh as “18 -> anon_inode:inotify” which makes sense, but may be is in issue for rakudo/moarvm 02:47
kalkin- (stupid reconnects) 02:48
timotimo how do you mean? in issue?
gotta head home and get some sleep 02:49
kalkin- timotimo: thanks for helping will experiment further 02:50
timotimo sure thing
kalkin- just using read(2) seems to work for me 03:05
BenGoldberg s: EVAL("1"), "Str", () 03:29
BenGoldberg peers around for saucebot... 03:30
samcv good * everyone. having been around the last day or two. 06:52
NeuralAnomaly, status
NeuralAnomaly samcv, [✘] Next release will be in 3 weeks and 6 days. Since last release, there are 1 new still-open tickets (1 unreviewed and 0 blockers) and 32 unreviewed commits. See perl6.fail/release/stats for details
samcv don't see the latest builds on rakudo.org 06:53
moritz \o 08:11
lol I blogged: perlgeek.de/blog-en/perl-6/2017-00...graph.html 08:12
Geth doc: titsuki++ created pull request #1203:
Add example for sigspace		 08:26
doc: 52960cc5d0 | titsuki++ | doc/Language/grammars.pod6
Add example for sigspace
doc: cef5f94dea | (Itsuki Toyota)++ | doc/Language/grammars.pod6
Merge pull request #1203 from titsuki/add-sigspace-example

Add example for sigspace
samcv sigspace is really weird in perl 6 08:31
m: say 'blah here' ~~ m:s/ blah here / 08:32
camelia ｢blah here｣
samcv m: say 'blah here ' ~~ m:s/ blah here /
camelia ｢blah here ｣
samcv both of these match, the last one matches the spaces but first one doesn't
m: say 'blah here ' ~~ m/ blah here /
camelia False
samcv m: say 'blah here ' ~~ / blah here /
camelia Nil
samcv m: say 'blah here ' ~~ m/ blah here /
camelia False
samcv m: say 'blah here ' ~~ :s/ blah here /
camelia 5===SORRY!5===
Regex not terminated.
at <tmp>:1
------> 3 'blah here ' ~~ :s/ blah here /7⏏5<EOL>
Regex not terminated.
at <tmp>:1
------> 3 'blah here ' ~~ :s/ blah here /7⏏5<EOL>
Unable to parse regex; coul…
samcv m: say 'blah here ' ~~ /:s blah here /
camelia ｢blah here ｣
samcv m: say 'blah here ' ~~ m/:s blah here / 08:33
camelia ｢blah here ｣
moritz m: say ' blah here ' ~~ m:s/ bla here / 08:37
camelia False
moritz m: say ' blah here ' ~~ m:s/ blah here /
camelia ｢blah here ｣
moritz m: say ' blah here ' ~~ m/ :s blah here /
camelia ｢blah here ｣
samcv it's basically like \s+ when you have a space when doing sigspace 08:38
well
m: say 'blah' ~~ /:s blah/
camelia ｢blah｣
samcv m: say 'blah' ~~ /:s blah /
camelia ｢blah｣
samcv that still matches though?
so it's like \s* at the last token but like \s+ on others? 08:39
m: say 'blah' ~~ /:s b lah /
camelia Nil
samcv m: say 'b lah' ~~ /:s b lah /
camelia ｢b lah｣
samcv m: say 'b lah ' ~~ /:s b lah /
camelia ｢b lah ｣
samcv m: say ' '.match: /<ws>/ 08:41
camelia ｢ ｣
ws => ｢ ｣
samcv m: say ' '.match: /<.ws>/
camelia ｢ ｣
samcv m: say ' '.match: / / 08:42
camelia 5===SORRY!5=== Error while compiling <tmp>
Null regex not allowed
at <tmp>:1
------> 3say ' '.match: / /7⏏5<EOL>
samcv m: say ' '.match: /' '/
camelia ｢ ｣
samcv m: say 'test '.match: /test<.ws>/
camelia ｢test ｣
samcv m: say 'test'.match: /test<.ws>/
camelia ｢test｣
samcv m: say 'test abc'.match: /test<.ws>/
camelia ｢test ｣
samcv m: say 'testabc'.match: /test<.ws>/ 08:43
camelia Nil
samcv so i guess end of string or start of string counts as whitespace
andrzejku hi 09:20
:)
parv unicorn is up
IOninja samcv: that's because it was a compiler release, not Star 09:28
samcv kk
parv, unicorn? 09:29
parv unicorn gundam (anime series; now "hunter x hunter" is on)
samcv ah
parv (in hindsight should have mentioned "gundam" instead) 09:30
RabidGravy Hmm 09:37
m: my $i = 0; say ((0 xx *).flatmap: { last if $i++ == 6; $i }).elems;
camelia Cannot .elems a lazy list
in block <unit> at <tmp> line 1

Actually thrown at:
in block <unit> at <tmp> line 1
RabidGravy is that a recent change? 09:37
IOninja RabidGravy: no, but flat was failing to preserve laziness and that was fixed recently. 09:38
IOninja m: my $i = 0; say ((0 xx *).flatmap: { last if $i++ == 6; $i }).eager.elems; 09:39
camelia 6
andrzejku guys what do you know about perl6 and xmpp? 09:41
RabidGravy there's a module for it that someone recently revived, that's about it 09:45
samcv i think i noticed that once IOninja. was it reifying it or just not preserving what is-lazy would give you? 09:46
RabidGravy IOninja, right that makes sense, so the code in DBIish that basically does that was working by accident then :)
IOninja samcv: not preserving .is-lazy, which caused things to reify stuff 09:49
samcv ah ok
IOninja m: my @a = (0 xx *).flat
camelia ( no output )
IOninja That used to hang
IOninja m: sub (*@a) { }((0 xx *).flat) 09:49
camelia ( no output )
IOninja that too 09:50
sena_kun does :dba adverb still alive? 10:00
moritz m: say 'bar' ~~ regex { :dba('foo') bar } 10:11
camelia Nil
moritz m: say 'bar' ~~ regex { bar }
camelia ｢bar｣
moritz huh
untested at least 10:12
Perl6::Grammar uses it quite a bit
sena_kun I'll file a roast issue now... 10:15
sena_kun moritz, thanks. 10:18
moritz QAST::Regex(:rxtype(literal) :subtype()) \"foo\" 10:24
m: say 'foobar' ~~ regex { :dba('foo') bar }
camelia ｢foobar｣
0 => ｢foo｣
sena_kun I'm working on tilde explanation now. It says that tilde usage will produce a better error message, but how I can see it(the error message) when regex failed? 10:29
m: say "abc" ~~ /a ~ (c) (b)/; say "ab" ~~ /a ~ (c) (b)/
camelia ｢abc｣
1 => ｢b｣
0 => ｢c｣
Nil
Geth DBIish: 37d5c2079c | (Jonathan Stowe)++ | lib/DBDish/StatementHandle.pm6
Make fetchall-* methods eager

The flatmap was accidentally reifying the lists until this was fixed in github.com/rakudo/rakudo/commit/ca...8c5da329c6
This is a naïve fix, but the tests would need to be rejigged if the
  laziness wants to be preserved.
Fixes #87
ZzZombo What is "braid" and why do I feel bad about it in advance? 10:42
masak ZzZombo: my answer will be only as far as I understand :)
a braid is the total sum of all languages and slangs currently in effect
masak like, say you just defined a new operator in your program. now you've technically extended the language with that operator. 10:43
that's like adding one little thread (your operator) to an already quite woven-together braid (Perl 6)
ZzZombo: the reason Perl 6 needs to invent a fairly new term for this is that it's trying to do something other languages are not (as far as I'm aware) 10:44
the thing it's trying is something like having the cake and eating it wrt everyone being able to extend the language while also minimally colliding with everyone else in doing so 10:45
ZzZombo well then
masak the closest thing in Perl 5 is something called "source filters". I know next to nothing about them, but I can say with certainty that they are not famous for their *composability* 10:46
I've gotten the eerie feeling that source filters are not very far removed from being regex-manipulations on source text -- someone feel free to enlighten me further on that point 10:47
hm, I wonder if someone has ever done something like source filters, but with PPI as a foundation :)
does PPI round-trip back to source text? 10:48
ZzZombo HELP! I'm here watching my cursor suddenly moving by itself towards the right bottom corner.... 10:50
I am not holding the mouse and its resting on its pad
it's greased, there is no way it can somehow start to move by itself... 10:51
masak how long since you performed the Mousa Calibration Ritual?
Mouse*
ZzZombo what is that? Is that some thing related to the old devices with a track ball inside them?
masak maybe you need to reboot your fridge or toaster of they're IoT devices. they might interfere with the mouse. 10:52
ZzZombo WTF, this is so creepy.
araraloren ... 10:53
masak maybe it's your neighbor who has hacked the mouse. you did password-protect your mouse, right? 10:56
ZzZombo hunter2
masak all I see is asterisks
ZzZombo see, it works.
masak oh, that was a password? great! 10:57
IOninja wtf 11:08
masak ZzZombo: I know TimToady and possibly jnthn have been making braid-related Rakudo commits lately, but I don't know what prompted it
IOninja RabidGravy: that code looks insane to me. What's with the 0 xx * bit? 11:09
RabidGravy dunno
IOninja RabidGravy: also, that's the wrong commit. The fix is in github.com/rakudo/rakudo/commit/51b0aba
masak `0 xx *` just means "a fair amount of zeros" :) 11:10
IOninja RabidGravy: and why eager the fetchall? You'd think that to be a heavy operation that can benefit greatly from not fetching everything at the same time 11:10
masak: ... which aren't used in that code.
RabidGravy hence "naive fix" 11:11
masak huh. 11:12
IOninja masak: my thoughts exactl.y
masak if you mean that they aren't used for the .map, that just means that the `0` could've been anything
IOninja like repeat...until? Sure. 11:13
masak basically.
RabidGravy I didn't write the code, I'm sure whoever did had their reasons
it wasn't lazy before due to the flat losing the laziness so I was just preserving the tested behaviour 11:16
IOninja It was lazy.
star: my $x = (0 xx *).flat; say $x[^10] 11:17
camelia (0 0 0 0 0 0 0 0 0 0)
RabidGravy well the tests all assumed it wasn't 11:18
IOninja :/ 11:22
Geth doc/tilde-explanation: 7453809342 | Altai-man++ | doc/Language/regexes.pod6
Remove information about :dba that is not in roast(now).
 11:37
RabidGravy m: say [(0 xx *)] eqv []; 11:50
camelia This type cannot unbox to a native integer: P6opaque, Failure
in block <unit> at <tmp> line 1
RabidGravy any value in making that a more obvious message? 11:51
ufobat .seen zostay 11:52
yoleaux I saw zostay 11 Dec 2016 18:10Z in #perl6: <zostay> No. Hard to find time and when I do I keep running into obscure async bugs
ufobat :/
Geth perl6.org: 514039ce2e | (Steve Mynott)++ | 576 files
Make it even more obvious we're reading an archive!
 11:58
IOninja s: &infix:<eqv>, \([(0 xx *)], [])
IOninja s: &infix:<eqv>, \([(0 xx *)], []) 11:59
SourceBaby IOninja, Sauce is at github.com/rakudo/rakudo/blob/b933...Mu.pm#L871
IOninja RabidGravy: I'd say there should not be any error there. 12:00
sena_kun stmuk_, ping? did you commit github.com/perl6/perl6.org/commit/514039ce2e because of mine question?
IOninja m: say (0 xx *) eqv (0 xx *) 12:01
camelia Cannot eqv lazy Sequences
in block <unit> at <tmp> line 1
stmuk_ if you are toddandmargo then yes :)
IOninja hm
RabidGravy IOninja, yeah I was just about to say that :)
IOninja ¯\_(ツ)_/¯ 12:02
m: say [(0 xx *)].is-lazy
camelia True
sena_kun stmuk_, ah, okay then. :)
RabidGravy and strangely I just typed that in the shell here too 12:03
sena_kun also, I double question above, since there are more people now. The tilde explanation(in specs, at least) says that "tilde usage will produce a better error message". but how I can see it(the error message) when regex failed? 12:04
m: say "abc" ~~ /a ~ (c) (b)/; say "ab" ~~ /a ~ (c) (b)/
camelia ｢abc｣
1 => ｢b｣
0 => ｢c｣
Nil
sena_kun m: say "abc" ~~ /a (b) (c)/; say "ab" ~~ /a (b) (c)/
camelia ｢abc｣
0 => ｢b｣
1 => ｢c｣
Nil
IOninja sena_kun: it's produced in the grammar 12:05
jnthn m: grammar G { token TOP { a ~ c b } }; G.parse('a b')
yoleaux 18 Feb 2017 16:42Z <IOninja> jnthn: seems $foo.WHAT === Code doesn't detect thunks? Here's my version of infix:<orelse> but it still returns a thunk instead of executed thunk. Am I doing it wrong? gist.github.com/zoffixznet/82ef0cf...5fab557427
camelia ( no output )
yoleaux 18 Feb 2017 16:53Z <IOninja> jnthn: nevermind, I've got Blocks not Code
IOninja m: grammar { token TOP { '[' ~ ']' .+ }; method FAILGOAL (|c) { say c; exit } }.parse: '[meow'
camelia \("']' ")
jnthn m: grammar G { token TOP { a ~ c b } }; G.parse('ab')
camelia ( no output )
IOninja m: grammar { token TOP { \[ ~ \] \w+ }; method FAILGOAL($goal) { die "cannot find $goal near position {self.pos}" }}.parse: '[moep bla]'
camelia cannot find \] near position 5
in method FAILGOAL at <tmp> line 1
in regex TOP at <tmp> line 1
in block <unit> at <tmp> line 1
jnthn Oh, right, you have to implement FAILGOAL
We used to generate this error by default 12:06
e.g. there was a default FAILGOAL that did "Cannot find ]" or so
andrzejku hi people :)
sena_kun jnthn, this default will be returned on some point or I can document it as "do it yourself"? 12:07
jnthn And pepole whined "oh but every other regex construct backtracks, why not this one", so we ended up changing it...
jnthn sena_kun: No, these days it's more like "you have the opportunity to give better errors by implementing FAILGOAL" 12:07
I suggest providing an example like the one IOninja just gave
sena_kun jnthn, FAILGOAL works only for ~ or is it a default method for failure errors in general? 12:08
jnthn sena_kun: Only for ~
sena_kun jnthn, noted. Thanks, folks, I'll finish it faster now. 12:09
jnthn sena_kun: The ~ is known as "goal matching syntax", which is where the name FAILGOAL came from
ufobat .tell zostay what is the status of your HTTP-Request-Supply? i'd really love to have it on modules.perl6.org so we can have a async http p6w server 12:15
yoleaux ufobat: I'll pass your message to zostay.
masak I think I was one of perhaps several who said "every other regex construct backtracks" :) 12:45
I remember arguing the point in an RT ticket somewhere 12:46
I think what stood out to me was that grammars already have a composable kind of failure mechanism (backtracking), and this one thing incongruously used exceptions instead 12:47
stmuk_ I think all the slangs in the ecosystem are broken now 12:48
is braid getting moved to 6.d? 12:49
pmurias stmuk_: I would be surprised if we support any backwards compability with slangs 12:55
sammers hi #perl6 13:01
DrForr Afternoon, sammers. 13:02
stmuk_ slangs aren't documented either so I guess its not really supported yet
sena_kun why is it broken? on latest rakudo mine just prints "Deprecated use of %*LANG<MAIN> assignment detected in use; module should export syntax using $*LANG.define_slang("MAIN",<grammar>,<actions>) instead". But it passess my tests. \0/ 13:07
sammers is there anyway to get the pid from Proc::Async? 13:12
RabidGravy isn't there a .pid attribute? 13:19
on Proc
sammers yes 13:20
but I am getting no such method for pid on Proc::Async 13:21
RabidGravy sorry, yes, confused myself there 13:22
IOninja don't really see a way to get at the Proc... from reading the code. 13:30
stmuk_: nope, it was merged yesterday. 13:32
IOninja %*LANG is not a supported way to do anything. We can't restrict our efforts just because someone went ahead and used some part of guts. 13:33
And the warning suggests a proper way to do what the slang modules are doing.
IOninja sena_kun: same with your warning. 13:34
sena_kun IOninja, yep, I understand it. But warning about deprecation is not "totally broken". 13:35
IOninja "If you have any code that depends on %*PRAGMAS or $*ACTIONS, it'll break."
"if you have code that sets %*LANG from a use, it'll emulate and warn"
sena_kun ah, okay.
IOninja "now that the release is out, I'm going to merge braids, so expect a bit of instability and/or warnings from slang modules"
IOninja and there's at least one slang module that's currently crashing with some weird error.... 13:36
DrForr Sigh, probably mine. 13:36
IOninja Slang::Tuxic 13:37
lizmat $ perl6 -MSlang::Tuxic -e 'say "hello world"'
hello world
not immediately obvious, or am I missing something 13:38
IOninja lizmat: [Tux] had issues with it: irclog.perlgeek.de/perl6-dev/2017-...i_14126519
lizmat is still slightly dazed from jetlag
sena_kun I wonder why I've been mentioned in the release announce. 13:41
IOninja "The following people contributed to this release:" 13:43
Everyone who commited to doc, roast, moarvm, nqp, or rakudo are listed.
ordered by number of commits 13:44
sena_kun Yep. But I did nothing in rakudo repo. Ah, so doc contributions are counted too. Cool.
IOninja Yeah. This is a script that finds contributors: github.com/rakudo/rakudo/blob/nom/...butors.pl6 13:45
sena_kun My self-confidence was incremented today. \0/ 13:46
IOninja \o/
tbrowder o/ salud, #perl6 13:50
sena_kun tbrowder, hello.
tbrowder hi, sena_kun
my question for today involves initializing and assigning to arrays 13:51
here's what i've tried 13:52
m: my %r<a> = Array.new; %r<a>.push: 'b' 13:53
camelia 5===SORRY!5=== Error while compiling <tmp>
Shaped variable declarations not yet implemented. Sorry.
at <tmp>:1
------> 3my %r<a>7⏏5 = Array.new; %r<a>.push: 'b'
tbrowder what is solution? 13:54
IOninja m: m: my %r; %r<a>.push: 'b'; say %r
camelia {a => [b]}
gfldex m: my %r = a => Array.new; %r<a>.push: 'b'; dd %r; 13:54
camelia Hash %r = {:a($["b"])}
DrForr m: my %r<a> = (); %r.push: 'a'; # if you want to be explicit.
camelia 5===SORRY!5=== Error while compiling <tmp>
Shaped variable declarations not yet implemented. Sorry.
at <tmp>:1
------> 3my %r<a>7⏏5 = (); %r.push: 'a'; # if you want to be
masak tbrowder: problem was the `my %r<a>` part
tbrowder: you probably meant `my %r; %r<a> = []` 13:55
DrForr Oh, ne'er mind, that's what I get for copying...
gfldex you can do it if you assign a Pair of Str and Array in the initialiser
tbrowder masak: i tried the [] before and it didn't work either, maybe my rakudo is too old 13:56
IOninja m: m: my %r; my $z = %r<a>; $z.push: 'b'; say %r
camelia {}
IOninja aww
m: m: my %r; my $z := %r<a>; $z.push: 'b'; say %r
camelia {a => [b]}
gfldex tbrowder: the syntax my %foo<key1 key2 key3> is reserved for Hashes with compile time keys.
IOninja hehee :) Magic!
m: my %r; my %z; my $z := %r<a>; my $r := $z; %z<z> := $r; $z.push: 'b'; say %r; say %z 13:57
camelia {a => [b]}
{z => [b]}		 13:58
IOninja ehehe way cool
hmmm
m: my %r; my $z; %r<a b c d e f g> := $z xx *; $z.push: 'b'; say %r;
camelia WARNINGS for <tmp>:
Useless use of $z in sink context (line 1)
Cannot bind to Hash slice
in block <unit> at <tmp> line 1
tbrowder woe is me, all i want to do is have a hash of arrays formed programmatically, but i guess masak's solution is good for now 13:59
IOninja what you taking aboot... it's not useless!
m: my %r; my $z; %r{$_} := $z for <a b c d e f g>; $z.push: 'b'; say %r;
camelia {a => [b], b => [b], c => [b], d => [b], e => [b], f => [b], g => [b]}
IOninja ehehe
That's pretty cool
DrForr m: my @foo = 1,2; my %r; %r<a> = @foo;
camelia ( no output )
IOninja tbrowder: you did see my version, right? You don't need any intializations. It'll autovivify automatically. 14:00
tbrowder i think so...let me look back
IOninja m: my %r; %r<a>.push: 42; %r<b>.append: 72, 89; %r<z> = [1, 2, 3, 4]; say %r 14:01
camelia {a => [42], b => [72 89], z => [1 2 3 4]}
IOninja The only suckage is:
m: my %r; say $r<foo>.elems
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$r' is not declared. Did you mean '%r'?
at <tmp>:1
------> 3my %r; say 7⏏5$r<foo>.elems
IOninja m: my %r; say %r<foo>.elems
camelia 1 14:02
IOninja m: my %r; %r<foo> andthen .elems.say
camelia ( no output )
tbrowder yes, i thought that was madak, sorry
IOninja m: my %r; %r<foo>:v.elems.say
camelia 0
tbrowder s/d/s/
IOninja m: my %r = foo => [<a b c>]; %r<foo>:v.elems.say
camelia 3
IOninja Oh, cool. No suckage then. 14:03
Geth doc/tilde-explanation: 49a1cef4fe | Altai-man++ | doc/Language/regexes.pod6
Add FAILGOAL explanation and example
tbrowder my %r; %r<a>.push: 'b' 14:04
m: my %r; %r<a>.push: 'b' 14:05
camelia ( no output )
sena_kun tbrowder, will you have some time for a review later?
tbrowder sena_kun: probably, just sitting around recovering from some winter crud 14:06
anyone have strong feelings on best p6 http client? 14:07
sena_kun ah, I spent apathetically a whole day yesterday. but didn't do any work. :/ 14:10
SmokeMachine moritz: I'm sorry, but I'm curious: why isn't Directory a child of File? 14:11
gfldex tbrowder: HTTP::Client does support https now, for small values of https that is.
Geth ecosystem: b93e3780f6 | (Ahmad M. Zawawi)++ | META.list
farabi6: META.info -> META6.json
sena_kun tbrowder, github.com/perl6/doc/pull/1190 this little peace and, more likely, github.com/perl6/doc/pull/1187, but the second is bigger(and older). Your help will be greatly appreciated. 14:12
tbrowder sena_kun: ok, it will be later today if that's ok 14:12
sena_kun tbrowder, any time you can and want is fine, it is me who is at fault here. :) 14:13
tbrowder gfldex: how small? i want a client that can handle com with acme server of letsencrypt 14:14
gfldex tbrowder: very little testing was done and don't expect man TLS related error messages. 14:15
*many
tbrowder well, i can try it...thanks
gfldex: probably the https certs are larger than the expected payload 14:17
SmokeMachine m: my %r; %r.push: a => 'b'; say %r 14:21
camelia {}
tbrowder m: my %r; %r<a>.push('a'); %r<a>.push('b'); my @a = %r<a>; say @a.perl 14:25
camelia [["a", "b"],]
tbrowder IOninja, it looks like i keep pushing to the first array element 14:27
tbrowder IOninja: looks like consecutive pushes are to the first array element. 14:29
m: my %r<a> = Array.new('a'); %r<a>.push: 'b'; say %r<a>.perl 14:32
camelia 5===SORRY!5=== Error while compiling <tmp>
Shaped variable declarations not yet implemented. Sorry.
at <tmp>:1
------> 3my %r<a>7⏏5 = Array.new('a'); %r<a>.push: 'b'; say
SmokeMachine m: my %r; %r.push: a => 'b'; say %r # docs.perl6.org/type/Hash#method_push 14:33
camelia {}
SmokeMachine m: my %r; %r.push: (a => 'b'); say %r
camelia {a => b}
SmokeMachine Why is the () needed? 14:34
m: my %r; push %r, a => 'b'; say %r
camelia {}
SmokeMachine m: my %r; push %r: a => 'b'; say %r 14:35
camelia {}
SmokeMachine m: my %h; push %h, 'd' => 5; dd %h # from the doc 14:36
camelia Hash %h = {:d(5)}
SmokeMachine m: my %h; push %h, d => 5; dd %h 14:36
camelia Hash %h = {}
tbrowder IOninja: i looked back at yr examples and still don't see a hash of arrays 14:37
SmokeMachine m: my %r; %r.push: 'a' => 'b'; say %r
camelia {a => b}
SmokeMachine Why it works with '' but doesn't work without it? 14:38
SmokeMachine m: my %r; %r.push: a => 'b'; m: my %h; %h.push: 'a' => 'b'; dd %r, %h 14:39
camelia Hash %r = {}
Hash %h = {:a("b")}
tbrowder my examples have been with one key, but the need is for a basic hash of arrays, one mutable array per hash key
how can that most painlessly and correctly be created incrementally? anyone? 14:40
in p5 terms i believe: an HoA 14:41
SmokeMachine m: dd a => 4; dd 'a' => 4 # why? 14:44
camelia block <unit>
:a(4)
gfldex m: role R { method push($key, $value){ self.{$key}.push: $value; self } }; my %h does R; %h.push('A', 1); %h.push('B', 2); %h.push('A', 3); dd %h; 14:46
camelia Hash+{R} %h = {:A($[1, 3]), :B($[2])}
gfldex tbrowder: ^^^
SmokeMachine gfldex: hash already has a push method, doesn't it? 14:48
gfldex SmokeMachine: yes
m: my %h; %h.push(a => 1); dd %h;
camelia Hash %h = {}
gfldex i did expect it to add the Pair to the Hash. 14:49
timotimo SmokeMachine: because a => 4 is a named parameter and 'a' => 4 is a pair literal that'll be put into a positional parameter
SmokeMachine m: my %h; %h.push('A' => 1); %h.push('B' => 2); %h.push('A' => 3); dd %h;
camelia Hash %h = {:A($[1, 3]), :B(2)}
SmokeMachine timotimo: ok, that makes sense... 14:50
gfldex timotimo: as stated in the docs :->
SmokeMachine gfldex: I didn't read everything... thanks... :( 14:51
timotimo yup, the docs say that
it's a common gotcha, especially for .push :) 14:52
gfldex luckily it wasn't me who wrote that
timotimo sooo, do the docs actually tell about s{abc} = "def"? i couldn't find anything about that form when i was answering that question on the users mailing list
IOninja recalls writing that
gfldex timotimo: it's not in the docs because it's not in roast 14:53
SmokeMachine Couldn't the hash's push receive named parameters and act as it was positional pairs? I'm not sure if it's a good idea...
IOninja timotimo: last code example: docs.perl6.org/language/operators#...bstitution
gfldex SmokeMachine: it could but then loads of ppl would not learn about the named-argument rule on Method 14:55
DrForr My mistake there, I'm afraid.
tbrowder ok, but why is, in my examples, %r<a>, not an array var?
SmokeMachine gfldex: thanks! 14:56
IOninja SmokeMachine: it'd make it much slower, lock us out from adding any options to the routine later on, and this quirk is present in literally all method calls, so adding support for that just in push is a drop in a bucket
tbrowder: which example exactly?
SmokeMachine IOninja: thanks!
timotimo OK, so there's a line in one of the examples... it's not very visible, though. like, it's not pointed out at all that that example is going to be presenting such a different syntax out of the blue 14:57
IOninja timotimo: well, it is, kinda. The last sentence is all about that syntax.
timotimo also: how the heck do we put it into the index? 14:58
i'd like it to have a heading of its own
tbrowder IOninja: my %r<a>; %r<a>.push: 'a'; # repeat push one or more times... 14:59
timotimo but "my %r<a>" doesn't even work?
tbrowder so no simple way to incrementally create a p6 HoA?, not good pr 15:00
timotimo of course there is? 15:01
let me try
Geth doc: 94a64bdbe3 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Method.pod6
index extra named arguments
doc: 1140358315 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Method.pod6
link to nextsame
timotimo m: my %h; %h<foo>.push(1); %h<foo>.push(2); %h<bar>.push(99); dd %h 15:02
camelia Hash %h = {:bar($[99]), :foo($[1, 2])}
timotimo what's wrong with this?
tbrowder but there needs at least needs to be a good example in the docs which i'm happy to add if i knew how to make one :)
IOninja tbrowder: well, you keep using my %r<a> even though it's been pointed out to you that's not correct syntax.
tbrowder: which one keeps pushing to the first element? 15:04
tbrowder sorry, but isn't that what timotimo just did?
timotimo no
i used "my %h"
pmurias n "my %r<a>;" is the part that does not work
timotimo "my %h<foo>" would (in the future, when we've implemented that) be a declaration for a hash that only accepts the key "foo" and will never accept any other key 15:05
tbrowder oh, slapping forehead,...gross series of typos
let me try again: 15:06
timotimo oh, i think i see what your problem is
m: my %h; %h<a>.push(1); %h<a>.push(2); my @a = %h<a>; my @b = @%h<a>; dd @a, @b;
camelia Array @a = [[1, 2],]
Array @b = [Failure.new(exception => X::AdHoc.new(payload => "Type Seq does not support associative indexing."), backtrace => Backtrace.new)]
timotimo whoops %)
m: my %h; %h<a>.push(1); %h<a>.push(2); my @a = %h<a>; my @b = @(%h<a>); dd @a, @b; 15:07
camelia Array @a = [[1, 2],]
Array @b = [1, 2]
timotimo tbrowder: your mistake was that the arrays in the hash are in an itemized container, so when you assigned to @a with "my @a = %h<a>" you were only putting a single item into @a
and you interpreted that as if %h<a> was a list with a list as its first element
tbrowder m: my %r; %r<a>.push: 'a'; %r<a>.push: '3'; say %r<a>.perl 15:08
camelia $["a", "3"]
timotimo exactly
tbrowder timotimo: thanks for yr patience, and i'll go back and ensure my prog uses right syntax, also thanks IOninja 15:11
timotimo sure thing
pmurias 'use Test; plan(1); ok(1)' just started working in local rakudo.js :) 15:22
ZzZombo # Sleep a bit to prevent 100% CPU usage
sleep 0.05;
Surely there is a better way to do that, isn't there? Like yield in some languages.
pmurias ZzZombo: where is that used? 15:23
IOninja heh 15:24
timotimo you could use NativeCall; sub pthread_yield is native(Str) returns int32 {*}; pthread_yield(); 15:25
IOninja And he's gone :) 15:27
ZzZombo github.com/azawawi/perl6-electron/...p.pm6#L120
IOninja Looks normal to me. 15:29
ZzZombo Speaking of that module, how does it work? Isn't Electron is JS+HTML+Node.js, so can I, say, populate items in a combobox dynamically from Perl using that, or what does that actually allow you to do? 15:30
IOninja You could use Supply.interval if you want it non-blocking.
or show the .run into start
*shove 15:31
Geth doc: f594c8e0bc | (Wenzel P. P. Peppmeyer)++ | doc/Type/Method.pod6
doc lastcall
 15:32
gfldex ENODOC-- 15:33
sadly I don't know what the current value of ENODOC is
pmurias ZzZombo: it seems to talk to electron using http 15:37
andrzejku hi, people 15:38
what do you know about Larry Wall and Perl6?
pmurias ZzZombo: the js backend I'm working on should allow to run on electron directly, but it a bit away from being usable as a toy and a long way away from being usable for something serious 15:39
s/to run/running
AlexDaniel andrzejku: ¯\_(ツ)_/¯ 15:40
ZzZombo What is Perl6 again? I forget.
AlexDaniel googles 15:41
“Larry Wall maintains a reference grammar known as STD.pm6”
:S
gfldex andrzejku: i know that the biggest value for „Larry Wall“ is TimToady and the biggest value for Perl 6 is Rakudo. 15:42
andrzejku :) 15:43
gfldex andrzejku: „biggest“ in this context means „largest impact on mankind“
El_Che andrzejku: you may confuse Larry Wall with that guy with the same name that succesfully started a business with colourfull shirts and cool hats: i.ytimg.com/vi/HAXGD3JE7pU/hqdefault.jpg 15:49
Geth doc: 86ff9ff317 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Hash.pod6
improve wording and examle
 15:56
ZzZombo Is there a way to output something like 1000/3 as such, but 5/10 as 0.5? 16:05
tbrowder timotimo: my examples today were very shortened from the code i'm working with. here today we could see the array form with the say .perl syntax, but in my real code i was seeing just a string concatenation of the array contents. i vaguely remember having the same problem last year but unpacking the array in code took some magic...don't know what best 16:07
practice is but i was creating a var like this: my @d = %r<a> which gave a single string but this works: my @d = @(%r<a>) which gave me the array elements
Geth doc: 31ac0d10c8 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Pair.pod6
show how to handle colon lists
 16:08
tbrowder looks kind of like dereferencing AND i failed to say the hash var was returned from the sub that actually created it!!
andrzejku what do you think if the people programming Perl are smarter than for ex. Ruby?
Perl6 suppose use more brain 16:09
tbrowder so i guess it is a reference, but i thought references were supposed to be kind of hidden for most uses, as in everything is kind of a reference 16:10
IOninja andrzejku: a nonsensical generalization.
sena_kun andrzejku, brainf-k supposes even heavier usage of brain activity. :)
IOninja m: .perl.say given <1000/3>, <5/10> 16:11
camelia (<1000/3>, 0.5)
IOninja ZzZombo: ^
m: .subst(:g, /<[<>]>/, '').perl.say given <1000/3>, <5/10> 16:12
camelia "333.333333 0.5"
IOninja m: .perl.subst(:g, /<[<>]>/, '').say given <1000/3>, <5/10>
camelia (1000/3, 0.5)
IOninja m: .perl.trans(«<>» => '').say given <1000/3>, <5/10> 16:13
camelia (1000/3, 0.5)
Geth doc: becaee574d | (Wenzel P. P. Peppmeyer)++ | doc/Type/Pair.pod6
index colon-list in Pair
IOninja tbrowder: no referencing or dereferencing involved. %r<a> array is itemized, so you're providing just 1 item to @a. The `@` is a coercer, not dereferencer. You could also slip with `| 16:17
m: my %r = :a[<a b c>]; my @d = |%r<a>; dd @d
camelia Array @d = ["a", "b", "c"]
IOninja m: my %r = :a[<a b c>]; my @d = @%r<a>; dd @d
camelia Array @d = [Failure.new(exception => X::AdHoc.new(payload => "Type Seq does not support associative indexing."), backtrace => Backtrace.new)]
IOninja m: my %r = :a[<a b c>]; my @d = @(%r<a>); dd @d
camelia Array @d = ["a", "b", "c"]
IOninja m: my %r = :a[<a b c>]; my @d = %r<a>.list; dd @d 16:18
camelia Array @d = ["a", "b", "c"]
IOninja m: my %r = :a[<a b c>]; my @d = %r<a><>; dd @d
camelia Array @d = ["a", "b", "c"]
tbrowder IOninja: thanks, that helps understand the array some better, i have to say the docs probaby could introduce arrays a little gentler, at least for me, haven't gotten head wrapped around itemization yet 16:21
IOninja If you have to boxes of crayons and I tell you to put one in the other, there are two ways to interpret it: shove the entire box into the other (so you retain information about the crayons inside of it) or move the crayons themselves, discarding one box. The former is itemization—treating something as one thing—the latter is the slipping of one array into another: you end up with one homogenous set of 16:24
elements.
*two boxes
m: my @box1 = <a b c>; my @box2 = <d e f>; @box1.push: @box2; say @box1 16:26
camelia [a b c [d e f]]
IOninja m: my @box1 = <a b c>; my @box2 = <d e f>; @box1.push: |@box2; say @box1
camelia [a b c d e f]
moritz of just use append
my @box1 = <a b c>; my @box2 = <d e f>; @box1.append: @box2; say @box1 16:27
m: my @box1 = <a b c>; my @box2 = <d e f>; @box1.append: @box2; say @box1
camelia [a b c d e f]
moritz pyhton does pretty much the same thing, excep that Perl 6's "push" is called "append" in python, and Perl 6's "append" is spelled += in python
tbrowder okay, i'll experiment some more then, thanks 16:33
timotimo there's also the "extend" method which maps to perl6's "append" 16:41
tbrowder sena_kun: pr 1190 looks good to me, too bad dba adverb isn't still usable!! 16:48
sena_kun tbrowder, it is usable(as I see it), just doesn't in roast yet. Thanks, I'll merge it right away. 16:49
*isn't
tbrowder hm, can you add a test in roast?
sena_kun tbrowder, hmmm, maybe I can, but not today, I think. I'm working on my course paper now. :/ Added this to my todo-list. 16:50
tbrowder ok, maybe just add an issue and say you will work on it. good luck on paper...subject? 16:52
Geth doc/master: 4 commits pushed by Altai-man++
tbrowder or title?
sena_kun tbrowder, I've opened an issue already. Title, eeeh... It is a rough translation from Russian(Ukrainian? Both), but "Usage of refinement types for a static verification of code invariants". Or something along the lines. 16:55
"of code properties" maybe fit better... anyway, not something interesting. 16:56
s/fit/fits/
Geth doc: 1509c4d1d6 | (Wenzel P. P. Peppmeyer)++ | 47 files
add tags to Language/*.pod6
 16:57
doc: 238865ae7e | (Wenzel P. P. Peppmeyer)++ | doc/Language/unicode_entry.pod6
forgot one
sena_kun -1 doc issue. only 216 left. \o/ 16:58
tbrowder sounds like a page turner!
sena_kun tbrowder, glad to hear it! 16:59
tbrowder sena_kun: the pr 1187 looks ok to me, too. have you figured out the conflict? 17:01
maybe try splitting into smaller prs
sena_kun tbrowder, it was my edits for output style, so I can fix it easily. The new web editor is very handy too. 17:02
Geth doc/master: 4 commits pushed by (Antonio Quinonez)++, Altai-man++ 17:04
stmuk_ odd 17:15
mouq.github.io/slangs.html 17:16
IOninja Odd what? 17:29
stmuk_ well the footer says its generated from a repo it isn't 17:33
I was git greping perl/doc but its a different repo 17:34
IOninja probably just a copy-paste of a page that used to be in docs but was taken down for Christmas because we never finalized slangs API 17:36
stmuk_ I don't think the content was actually ever in the docs repo itself but generally yes 17:42
timotimo no, you misread the footer 17:51
"using htmlify.p6 from perl6/doc on github"
that's the intended bracketing for that phrase 17:52
regnarg Plans for a Perl6 chromium-based web browser: github.com/regnarg/perl6-cef Currently it doesn't do much apart from initializing the CEF using NativeCall but it's a start... 17:53
IOninja cool 17:54
travis-ci Doc build errored. Altai-man 'Merge branch 'master' into grammar' 17:54
travis-ci.org/antquinonez/doc/builds/203190820 github.com/antquinonez/doc/compare...93f4d93261
regnarg It might need some improvements to NativeCall, I think it's currently not possible to put function pointers to structs, which half of the CEF API is comprised of. But some callback code is in NativeCall already and shouldn't be hard to adapt. 17:56
timotimo i think you can nativecast a pointer to a function by giving a signature or something like that 17:58
regnarg I will more often need the opposite: put a pointer to a Perl6 function (callback) into a struct. I'm not sure whether it can be currently done with some casting because it requires generating a C wrapper function. But the required infrastructure is already there because you can already pass Perl6 callbacks as arguments to C functions (at least the docs say so). 18:00
timotimo nativecall does the c wrapper thing for you
AlexDaniel m: my @manna = :a1, :b2, :c3; dd @manna 18:01
camelia Array @manna = [:a1, :b2, :c3]
AlexDaniel m: my @manna = :a1 :b2 :c3; dd @manna
camelia Array @manna = [:a1]
AlexDaniel hm
timotimo you might have to create a function that takes a pointer and immediately returns it to appease nativecall?
RabidGravy I think there is support for callable attributes in CSTructs
IOninja There's a ticket for that.
regnarg That might run into problems with the lifetime of the C wrapper
AlexDaniel regnarg: nice! 18:02
regnarg It might get freed after the function returns
Specifically you cannot just keep a void pointer around because NativeCall would not know when to free the wrapper
So my guess is that some (possibly simple) changes to NativeCall would be needed. 18:03
But I might be wrong, haven't yet looked into the details.
timotimo no, the wrappers leak
IOninja AlexDaniel: rt.perl.org/Ticket/Display.html?id...et-history and rt.perl.org/Ticket/Display.html?id...et-history and rt.perl.org/Ticket/Display.html?id...et-history
timotimo they get freed at end of life at the moment
IOninja more or less
regnarg Ok, in that case it might just happen to work
Even though it's probably not a robust solution in the long term 18:04
timotimo right, something clever is needed
regnarg I'll try that sometime soon
timotimo you want to just pass an anonymous block to a c function and not worry about it
AlexDaniel IOninja: so add RT #130816 to the list? :)
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130816
timotimo but you'll have to also assign it to a variable and manually free it at some point with some kind of API that we'll have to design
AlexDaniel not sure if I will ever be able to use chromium-based anything because it never respected my system settings, but this looks like a great project :) 18:05
regnarg Chromium has a very good core (w.r.t. speed and security) but it UI and especially configurability is horrendous. 18:06
But the CEF API is quite flexible (you can intercept every key event, every request, and so on) so one should be able to build something more flexible atop it. 18:07
robertle i have some functionality i want to put in a module, but I sometimes need it exposed in an OO fashion, sometimes procedurally. I could put both in the same module, but is that a terrible idea? so you would "use MuStuff" which would give you a role "MyStuff" as well as a sub my-stuff 18:15
robertle alternatively these could be two modules, e.g. Role::MyStuff and Module::MyStuff... not sure what is worse 18:15
DrForr I'd do the export module and have it be just a wrapper to methods in the main module. 18:17
RabidGravy yeah 18:20
robertle so two modules, the OO one wrapping thje procedural one?
regnarg Alternatively, you can use the export groups functionality 18:23
Or sets or how it is called
docs.perl6.org/language/modules#Ex..._Importing 18:24
and then the user does either use MyModule :oop; or use MyModule :procedural; or even use MyModule :ALL; 18:25
It might be cleaner than two modules but it depends on the specifics
robertle oh wow, that sounds great I'll do that!
specifics: it's like "Pluggable" 18:26
Geth doc: fdf275abd2 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Pair.pod6
make index entry user friendly
 18:31
pmurias is there a reason for IO::Handle.write to return True rather then the amount of bytes written? 18:58
AlexDaniel IOninja: ↑ 19:07
:)
Geth doc: b8af670621 | (Wenzel P. P. Peppmeyer)++ | doc/Language/traps.pod6
show constant fat arrow trap
 19:09
Geth doc: 84fbe82117 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Bag.pod6
link to Associative
 19:17
Geth doc: 4afe7dd9b5 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
doc Any.Bag and Any.BagHash
 19:30
doc: c95320f986 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Language/traps.pod6
“it's” → “its”
 19:32
regnarg ^ Language traps indeed ;-) 19:34
Geth doc: dbfa3fbdf9 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
doc Any.Set|SetHash|Mix|MixHash
 19:40
doc: f4f9d7b22d | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
doc Any.Supply
gfldex i just checked Any.pod6 with util/list-missing-methods.p6 and decided to look the other way 19:43
timotimo Any has all the methods 19:45
moritz we should really have called it All, not Any :-) 19:46
gfldex wait! I know Perl 6 so I can write a script that will generate all those docs! 19:48
moritz maybe we could have scheme where for example Str methods that appear in Any are actually documented in Str, and there's an :inject<Any> somwhere that copies the docs to Any 19:52
and some magic in htmlify to make that happen
Geth doc: 80d01ab35b | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
doc Any.Array|List|Map
 19:55
gfldex moritz: the method will show up on the page, just a little further down. And the information where the method actually comes from may be important to the reader. 19:57
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'link to Associative' 20:08
travis-ci.org/perl6/doc/builds/203219202 github.com/perl6/doc/compare/b8af6...fbe8211791
Geth perl6-most-wanted: e4c5587f7d | (Tom Browder)++ | most-wanted/modules.md
the p5 module is a good model
 20:08
doc: ac7659bb01 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
fix Any.list|push
Geth doc: d2cc944053 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
doc Any.Hash|Slip|hash
 20:11
Geth doc: 8372fe9e78 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
fix Defined as:
 20:20
Geth doc: bdf8ca8cb2 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
add methods that are TODO for Any
 20:27
SmokeMachine Is camelia going to run rakudo on js? 21:06
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'doc Any.Hash|Slip|hash' 21:16
travis-ci.org/perl6/doc/builds/203231004 github.com/perl6/doc/compare/ac765...cc94405337
travis-ci Doc build errored. Wenzel P. P. Peppmeyer 'fix Defined as:' 21:29
travis-ci.org/perl6/doc/builds/203232955 github.com/perl6/doc/compare/d2cc9...72fe9e78c9
Geth doc: 3085d7c9b6 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod6
doc Any.min|max|minmax
 22:04
BenGoldberg m: use NativeCall; dd cglobal('libc.so.6', 'errno', int32) 22:49
camelia 2
BenGoldberg m: use NativeCall; dd cglobal('libc.so.6', 'nosuchvar', int32)
camelia Cannot locate symbol 'nosuchvar' in native library 'libc.so.6'
in block at /home/camelia/rakudo-m-inst-1/share/perl6/sources/24DD121B5B4774C04A7084827BFAD92199756E03 (NativeCall) line 422
in block <unit> at <tmp> line 1
BenGoldberg m: use NativeCall; dd cglobal('libc.so.6', 'errno', int) 23:03
camelia 2
tbrowder ref run and shell: the docs say something like to use shell in lieu of run if you need special redirection or chars that need special interp by the shell. but, since i started with p6 it always seems that run can't handle. much of anything 23:04
tbrowder m: run "ps -C http o cmd=" 23:05
camelia run is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in sub run at src/RESTRICTED.setting line 14
in block <unit> at <tmp> line 1
gfldex tbrowder: you need to separate arguments by hand for run 23:06
IOninja run <ps -C http o cmd=> 23:07
AlexDaniel
.oO( run ‘ps’, ‘-C’, ‘http’, ‘o’, ‘cmd=>’ # no magic required )
 23:08
tbrowder hm, well, i'll try that, but the other issue is that the docs don't mention that you can use a shell invocation just like run to create a new Proc object 23:09
AlexDaniel “The return value is of type Proc.”
docs.perl6.org/routine/shell 23:10
tbrowder yes, it does say that, but i think it also should show creating a new proc object just as run. any objections to adding it? 23:11
tbrowder also i have used run with just 2 args: 'cmd', 'cmd args' 23:14
Geth doc: 4a33f5e852 | (Wenzel P. P. Peppmeyer)++ | doc/Type/IO.pod6
provide example for run
 23:15
BenGoldberg dd print => 1
tbrowder i'll go back and make sure i didn't make another dumb error
BenGoldberg m: dd print => 1
camelia block <unit>
BenGoldberg m: dd 'print' => 1
camelia :print(1)
IOninja m: IO::Handle.new(:path<STDIN>) 23:18
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of <STDIN>; in Perl 6 please use $*IN.lines (or add whitespace to suppress warning)
at <tmp>:1
------> 3IO::Handle.new(:path<7⏏5STDIN>)
IOninja Really? -_- 23:19
timotimo that's a bad miss :D 23:26
tbrowder AlexDaniel: you are right, every little piece! no wonder i have so much trouble with run. 23:35
tbrowder i really think the docs need tweaking to favor shell for casual use--why should i want to break up complex shell commands piece by piece? got a good use case? 23:37
gfldex i just broke the highlighter :) docs.perl6.org/type/IO#sub_run 23:39
AlexDaniel tbrowder: no.
gfldex tbrowder: it's the other way around, use run for casual use and be really damn careful with shell 23:40
tbrowder: shell escapes are tricky
AlexDaniel tbrowder: you can read this blog post if you are interested: github.com/perl6/specs/pull/102 23:41
Ulti so whats the solution to avoiding %*LANG 23:51
tbrowder AD, i have read that blog post... 23:53
tbrowder btw, i can't remember ever using escapes in a shell command, ever, in or out of perl 5 or 6. btw, glad to see p6 mkdir defaults to the -p arg that p5 didn't last time i looked. that was one thing that always made me go to shell. 23:57
AlexDaniel well, unless your path is fixed, you'll have a security hole right there… 23:59

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!