»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.
viki "Although the above example works as intended the eqv code has to fall back to a slower code path in in order to do its job." docs.perl6.org/routine/eqv 00:02
viki isn't sure it's a good idea to tell users so much about internals of compiler X, since the docs are meant to be compiler-agnostic
00:06 canopus joined 00:07 Ven left 00:23 adu joined 00:25 bjz left 00:26 Actualeyes left 00:32 perlawhirl left 00:33 ponbiki left 00:36 skaji_ joined 00:41 Actualeyes joined, Xliff joined
Xliff m: constant T = ( a=>1, b=>2); say T<a>; 00:42
camelia rakudo-moar e0c0ae: OUTPUT«Type List does not support associative indexing.␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤»
Xliff m: constant T = { a=>1, b=>2 }; say T<a>;
camelia rakudo-moar e0c0ae: OUTPUT«1␤»
viki There's a ticket for that 00:44
you can also use %() instead of {} 00:45
Xliff Ah. Thanks. 00:46
dalek c: fa0b8f6 | (Zoffix Znet)++ | doc/Language/quoting.pod6:
Differentiate <...> and qw|...|

Fixes RT#130184: rt.perl.org/Ticket/Display.html?id=130184		 00:47
synopsebot6 Link: doc.perl6.org/language/quoting
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130184
00:47 bjz joined
Xliff m: $a = 'b'; given $a { when 'b' || 'bb' || 'bbb' { say 'B'; } default { say 0; } } 00:48
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$a' is not declared␤at <tmp>:1␤------> 3<BOL>7⏏5$a = 'b'; given $a { when 'b' || 'bb' ||␤»
Xliff m: my $a = 'b'; given $a { when 'b' || 'bb' || 'bbb' { say 'B'; } default { say 0; } }
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Strange text after block (missing semicolon or comma?)␤at <tmp>:1␤------> 3{ when 'b' || 'bb' || 'bbb' { say 'B'; }7⏏5 default { say 0; } }␤»
Xliff m: my $a = 'b'; given $a { when 'b' || 'bb' || 'bbb' { say 'B'; }; default { say 0; } }
camelia rakudo-moar e0c0ae: OUTPUT«B␤»
Xliff m: my $a = 'b1'; given $a { when 'b' || 'bb' || 'bbb' { say 'B'; }; default { say 0; } }
camelia rakudo-moar e0c0ae: OUTPUT«0␤»
Xliff m: my $a = 'c1'; given $a { when 'b' || /^^c/ { say 'BC'; }; default { say 0; } } 00:49
camelia rakudo-moar e0c0ae: OUTPUT«0␤»
Xliff m: my $a = 'c1'; given $a { when 'b' || /^c/ { say 'BC'; }; default { say 0; } }
camelia rakudo-moar e0c0ae: OUTPUT«0␤»
Xliff huh
dalek c: c94df2f | (Zoffix Znet)++ | doc/Language/glossary.pod6:
Correct allomorphs section to include Rat and Complex literals behaviour
 00:56
synopsebot6 Link: doc.perl6.org/language/glossary
00:57 pierre_ joined, pierre_ left
viki m: my $a = 'c1'; given $a { when 'b' | /^c/ { say 'BC'; }; default { say 0; } } 00:57
camelia rakudo-moar e0c0ae: OUTPUT«BC␤»
viki Xliff: it uses smartmatch
Xliff Is that the difference between | and ||? 00:58
Thanks.
viki Xliff: || is higher precedence `or` and | constructs an any Junction 00:59
huggable: Junction
huggable viki, Logical superposition of values: docs.perl6.org/type/Junction
viki m: $_ = 'c42'; say 'c1' ~~ ('b' || /^c/) 01:00
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
viki Not really sure what that means TBH
Ohhh
Xliff: it turns into just 'b': you're with 'b' || /^c/, call ACCEPTS 01:01
m: say 'c1' ~~ (False || /^c/)
camelia rakudo-moar e0c0ae: OUTPUT«｢c｣␤»
viki yup
RT#130184 is pretty interesting. The docs never even say that <a b 42> would eqv ('a', 'b', '42'). Someone just assumed it implied that and now there's a StackOverflow page with a wall of text and half the answers claiming we got a bug in <...> :) 01:10
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130184
dalek c: 15aff17 | (Zoffix Znet)++ | doc/Language/quoting.pod6:
Fix broken L<>
 01:15
synopsebot6 Link: doc.perl6.org/language/quoting
01:17 perlawhirl joined 01:19 aborazmeh joined, aborazmeh left, aborazmeh joined
dalek osystem: 454684a | 0racle++ | META.list:
Add Pod::To::Markdown::Fenced
 01:19
osystem: 2b5486b | 0racle++ | META.list:
Merge pull request #269 from 0racle/patch-2

Add Pod::To::Markdown::Fenced
c: 82f7e01 | (Zoffix Znet)++ | doc/Language/glossary.pod6:
Expand allomorph glossary

  - Explain more what "two values" means
  - Mention that Setties and Baggies care about object identity
synopsebot6 Link: doc.perl6.org/language/glossary
AlexDaniel viki: that's not the first time I see people confused because of this 01:22
but it's not only the documentation, it is indeed confusing
viki I think the confusion lies in allomorphs being described as something magical that turns into an Int or a Str magically.... when it's just a subclass of both Str and corresponding numeric 01:23
I mean, you wouldn't expect class Meows { has $.value = 'meows' }.new added into a Bag to count as same thing as 'meows', even through they would print the same. 01:24
*class Meows is Str { 01:25
AlexDaniel sure but you're like
m: say 3 ∈ < 1 2 3 4 >
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
AlexDaniel no, doesn't work
m: say ‘3’ ∈ < 1 2 3 4 >
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
AlexDaniel no, doesn't work!
viki m: say class Meows { has $.value = 'meows' }.new eq 'meows'
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
viki m: say class Meows is Str { has $.value = 'meows' }.new eq 'meows'
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
viki :(
eh, whatever 01:26
AlexDaniel I'm not sure how would I explain it to a begginer if I had to
viki AlexDaniel: but in that example you used a Set operator on two items that aren't even sets. 01:27
AlexDaniel m: say <3> ∈ < 1 2 3 4 >
camelia rakudo-moar e0c0ae: OUTPUT«True␤»
AlexDaniel good point though 01:28
viki m: say <3/4> ∈ < 1 2 3/4 4 >; # ha-ha
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
viki :)
01:29 mr_ron joined
AlexDaniel m: say < 3/4 > ∈ < 1 2 3/4 4 >; # ha-ha 01:29
camelia rakudo-moar e0c0ae: OUTPUT«True␤»
AlexDaniel yea…
viki ^_^ 01:30
same with Complex
AlexDaniel huh?
TimToady the literal form doesn't admit whitespace 01:31
AlexDaniel ahhhhhhhhhhhhhhhhh yeaahhhhhhh…
m: dd <3i>
camelia rakudo-moar e0c0ae: OUTPUT«ComplexStr.new(<0+3i>, "3i")␤»
AlexDaniel m: dd <1+3i>
camelia rakudo-moar e0c0ae: OUTPUT«<1+3i>␤»
TimToady and requires both parts, apparently :) 01:32
AlexDaniel I wish there was a good way to cover all that
viki points to Zoffix's latest few commits to doc repo
TimToady wonders how soon till Zoffix disowns viki :) 01:33
AlexDaniel viki: I am not sure that any amount of docs will do, but that's a good try, yeah
viki <...> gives you allomorphs. You can also use it to construct literal Rats and Complex by wrapping it around what would normally be a division or addition operation (<42/10> or C<42+5i>) 01:35
<3i> gives you an allomorph, 'cause there's no addition involved, while <1+3i> gives you a Complex 'cause there is an addition 01:36
AlexDaniel m: dd <Inf> 01:37
camelia rakudo-moar e0c0ae: OUTPUT«NumStr.new(Inf, "Inf")␤»
AlexDaniel m: dd <∞>
camelia rakudo-moar e0c0ae: OUTPUT«"∞"␤»
viki ehehe
m: dd <NaN>
camelia rakudo-moar e0c0ae: OUTPUT«NumStr.new(NaN, "NaN")␤»
viki neat
AlexDaniel I mean, wtf is that? Is it a bug or?
viki m: dd <−42
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in quote words; couldn't find final '>'␤ ␤at <tmp>:1␤------> 3dd <−427⏏5<EOL>␤ expecting any of:␤ argument list␤ quote words␤ t…»
viki m: dd <−42> 01:38
camelia rakudo-moar e0c0ae: OUTPUT«IntStr.new(-42, "−42")␤»
viki AlexDaniel: yeah, I'd call that a bug
AlexDaniel /o\
viki :)
AlexDaniel m: dd <−42>
camelia rakudo-moar e0c0ae: OUTPUT«IntStr.new(-42, "−42")␤»
AlexDaniel u: −−
unicodable6 AlexDaniel, U+2212 MINUS SIGN [Sm] (−)
viki m: dd <−4/2>
camelia rakudo-moar e0c0ae: OUTPUT«RatStr.new(-2.0, "−4/2")␤»
AlexDaniel ah that's what you've tried too, ok
viki And that too
But that's already part of the thing that I was doing with -Inf and stuff 01:39
AlexDaniel that too?
viki m: dd <−4/2>
camelia rakudo-moar e0c0ae: OUTPUT«RatStr.new(-2.0, "−4/2")␤»
viki m: dd <-4/2>
camelia rakudo-moar e0c0ae: OUTPUT«-2.0␤»
AlexDaniel ah dammit
/o\
viki |
/\
AlexDaniel personally, I try to avoid <> 01:40
that thing is unpredictable
viki Like a loaded gun? :D
AlexDaniel loaded footgun, yes
viki m: dd Version.new: <I ♥ Perl 6>, 'I love Perl 6!' 01:50
camelia rakudo-moar e0c0ae: OUTPUT«Version.new('I love Perl 6!')␤»
viki We have this multi for constructing a Version object, but it appears its .perl is wrong 01:51
(oh and this multi isn't documented)
01:52 finanalyst joined
viki And I'm saying this 'cause I'm too lazy to do anything about those two things ^_^ 01:52
m: dd "77\x[309]7".comb(/:r '*' || \d+ <?{ $/.chars == $/.codes }> || <.alpha>+/).eager 01:55
camelia rakudo-moar e0c0ae: OUTPUT«slip()␤»
viki slip() ?? :/
Ah 01:57
\d+ matches the entire string, and then fails and there's no backtracking
AlexDaniel m: dd "77\x[309]7".comb(/:r ‘*’ || \d+ { say $/} <?{ $/.chars == $/.codes }> || <.alpha>+/).eager 02:00
camelia rakudo-moar e0c0ae: OUTPUT«｢77̉7｣␤｢77̉7｣␤｢77̉7｣␤slip()␤»
viki weird innit? 02:01
m: dd "77\x[309]7".comb(/:r ‘*’ || \d+ { say $/} <?{ $/.chars == $/.codes }> || <.alpha>+ || 'something else'/).eager
camelia rakudo-moar e0c0ae: OUTPUT«｢77̉7｣␤｢77̉7｣␤｢77̉7｣␤slip()␤»
02:02 Zoffix joined
Zoffix stupid host died! 02:02
02:02 Eddward joined
Zoffix m: dd "777\x[309]7".comb(/:r ‘*’ || \d+ { say $/} <?{ $/.chars == $/.codes }> || <.alpha>+ || 'something else'/).eager 02:02
camelia rakudo-moar e0c0ae: OUTPUT«｢777̉7｣␤｢777̉7｣␤｢777̉7｣␤｢777̉7｣␤slip()␤»
Zoffix m: dd "77777777\x[309]7".comb(/:r ‘*’ || \d+ { say $/} <?{ $/.chars == $/.codes }> || <.alpha>+ || 'something else'/).eager
camelia rakudo-moar e0c0ae: OUTPUT«｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤｢77777777̉7｣␤slip()␤»
Zoffix wtf?
viki . 02:03
AlexDaniel m: dd "77\x[309]7".comb(/:r \d+ { say $/ } <?{ False }> /).eager
camelia rakudo-moar e0c0ae: OUTPUT«｢77̉7｣␤｢77̉7｣␤｢77̉7｣␤slip()␤»
viki Zoffix: shoo. you're blowing my cover
02:03 Zoffix left
AlexDaniel ok let's start with something simple… 02:04
m: dd ‘777’.comb(/ :r .+ <?{ say $/; False }> /)
camelia rakudo-moar e0c0ae: OUTPUT«｢777｣␤｢777｣␤｢777｣␤slip()␤»
viki m: dd ‘777’.comb(/ .+ $ <?{ say $/; False }> /) 02:05
camelia rakudo-moar e0c0ae: OUTPUT«｢777｣␤｢777｣␤｢777｣␤slip()␤» 02:06
AlexDaniel or maybe even this
m: dd ‘789’.comb(/ . <?{ say $/; False }> /)
camelia rakudo-moar e0c0ae: OUTPUT«｢7｣␤｢7｣␤｢7｣␤slip()␤»
viki m: dd ‘777’ ~~ / .+ $ <?{ say $/; False }> /
camelia rakudo-moar e0c0ae: OUTPUT«｢777｣␤｢777｣␤｢777｣␤Nil␤»
viki m: dd ‘777’ ~~ / . $ <?{ say $/; False }> /
camelia rakudo-moar e0c0ae: OUTPUT«｢7｣␤Nil␤»
AlexDaniel m: dd ‘789’.comb(/ (.) <?{ say $0; False }> /)
camelia rakudo-moar e0c0ae: OUTPUT«｢7｣␤｢8｣␤｢9｣␤slip()␤»
AlexDaniel ok so you should be doing $0.chars == $0.codes I think 02:07
viki AlexDaniel: that one isn't very good, 'cause it's a comb
m: dd ‘777’ ~~ / .+ $ <?{ say $/; False }> /
camelia rakudo-moar e0c0ae: OUTPUT«｢777｣␤｢777｣␤｢777｣␤Nil␤»
viki m: dd ‘777’ ~~ / (.+) $ <?{ say $0; False }> /
camelia rakudo-moar e0c0ae: OUTPUT«｢777｣␤｢77｣␤｢7｣␤Nil␤»
AlexDaniel viki: to be honest I do not understand a thing here
Eddward Is there a such thing as mutable junctions? all($x, $y, $z) = "default"; 02:08
viki Eddward: what would that do with, say a one() junction? 02:09
or none()
Eddward Good question. I only considered all & any.
viki AlexDaniel: I think I kinda sorta get it. It matches something, gets to the block, fails, tries some more matching, etc 02:10
m: dd ‘777’ ~~ /:r (.+) $ <?{ say $0; False }> /
camelia rakudo-moar e0c0ae: OUTPUT«｢777｣␤｢77｣␤｢7｣␤Nil␤»
Eddward one could pick one at random to set, but that's want I was thinking any would do. Dunno what it should do now.
viki I don't fully get "backtracking" and stuff
Eddward none would be pointless.
AlexDaniel m: dd ‘789’ ~~ /:r (.+) $ <?{ say $0; False }> / 02:11
camelia rakudo-moar e0c0ae: OUTPUT«｢789｣␤｢89｣␤｢9｣␤Nil␤»
AlexDaniel viki: but why would $/ return a full string?
viki m: sub assign-stuff($where, $what) { $where = $what }; my ($x, $y, $z); assign-stuff all($x, $y, $z), "default"; dd [ $x, $y, $z ]
camelia rakudo-moar e0c0ae: OUTPUT«Cannot assign to a readonly variable or a value␤ in sub assign-stuff at <tmp> line 1␤ in sub assign-stuff at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
viki m: sub assign-stuff($where is rw, $what) { $where = $what }; my ($x, $y, $z); assign-stuff all($x, $y, $z), "default"; dd [ $x, $y, $z ]
camelia rakudo-moar e0c0ae: OUTPUT«["default", "default", "default"]␤»
viki m: sub assign-stuff($where is rw, $what) { $where = $what }; my ($x, $y, $z); assign-stuff any($x, $y, $z), "default"; dd [ $x, $y, $z ] 02:12
camelia rakudo-moar e0c0ae: OUTPUT«["default", "default", "default"]␤»
viki m: sub assign-stuff($where is rw, $what) { $where = $what }; my ($x, $y, $z); assign-stuff none($x, $y, $z), "default"; dd [ $x, $y, $z ]
camelia rakudo-moar e0c0ae: OUTPUT«["default", "default", "default"]␤»
viki heh
AlexDaniel Eddward: what are you actually trying to do?
viki AlexDaniel: no idea
(about the $/ whole thing)
AlexDaniel commit: all dd ‘789’.comb(/ . <?{ say $/; False }> /) # just in case 02:13
ah dammit
02:13 bjz_ joined
AlexDaniel commit: all dd '789'.comb(/ . <?{ say $/; False }> /) # just in case 02:13
viki :D
committable6 AlexDaniel, gist.github.com/d8136d4a67c79e9ad1...1ee114e943
Eddward AlexDaniel: I'm still playing with the idea of setting all the items in a slice to the same value.
committable6 AlexDaniel, gist.github.com/5b2b132079fbad410a...fa24c765a8 02:14
AlexDaniel Eddward: isn't ‘foo’ xx * enough?
viki "value" xx * is as good as it gets IMO
both as speed and readability go
Eddward I'm not sure I really like %<a b c> = 3 xx *
AlexDaniel viki: now wtf is that?
viki Where?
Eddward A typo
AlexDaniel commit: 2016.10,2016.11 dd '789'.comb(/ . <?{ say $/; False }> /) 02:15
committable6 AlexDaniel, ¦«2016.10»: ｢7｣␤｢7｣␤｢7｣␤().Seq␤¦«2016.11»: ｢7｣␤｢7｣␤｢7｣␤slip()
Eddward In my sample. ... and I replied to the wrong person. Sorry
viki :o
02:16 bjz left, finanalyst left
AlexDaniel bisect: old=2016.10 new=2016.11 dd '789'.comb(/ . <?{ say $/; False }> /) 02:16
viki Well, slip() is Empty
bisectable6 AlexDaniel, Bisecting by output (old=2016.10 new=2016.11) because on both starting points the exit code is 0
AlexDaniel let's see
bisectable6 AlexDaniel, bisect log: gist.github.com/20a7fd59aa70a01e63...b9353be607
AlexDaniel, (2016-10-24) github.com/rakudo/rakudo/commit/17...dac2b8f529
AlexDaniel okay
viki heh 02:17
AlexDaniel well, this doesn't look good
viki m: dd '123'.comb.^name
camelia rakudo-moar e0c0ae: OUTPUT«"Seq"␤»
viki And the docs say it returns a List :)
AlexDaniel /o\
viki m: my @a = '123'.comb('a'); dd @a.chars 02:18
camelia rakudo-moar e0c0ae: OUTPUT«0␤»
viki star: my @a = '123'.comb('a'); dd @a.chars
camelia star-m 2016.10: OUTPUT«0␤»
AlexDaniel m: dd '789'.comb(/ . <?{ say $/; True }> /)
camelia rakudo-moar e0c0ae: OUTPUT«｢7｣␤｢8｣␤｢9｣␤("7", "8", "9")␤»
AlexDaniel it does!
viki it? does? it who does what?
AlexDaniel m: say WHAT ‘789’.comb(/ . /) 02:19
camelia rakudo-moar e0c0ae: OUTPUT«(List)␤»
AlexDaniel that's a list!
m: say WHAT ‘789’.comb(/ . <!> /)
camelia rakudo-moar e0c0ae: OUTPUT«(Slip)␤»
AlexDaniel star: say WHAT ‘789’.comb(/ . <!> /)
camelia star-m 2016.10: OUTPUT«(Seq)␤»
viki dafuq
star: say WHAT ‘789’.comb(/ . /) 02:20
camelia star-m 2016.10: OUTPUT«(Seq)␤»
viki hehe
AlexDaniel ooooooooooooooh
viki Well, some of this may be intetional actually
AlexDaniel commit: 1794328c^,1794328c say WHAT ‘789’.comb(/ . /)
committable6 AlexDaniel, ¦«1794328c^»: (Seq)␤¦«1794328»: (List)
AlexDaniel aaaalright
viki I recall something or other about .match caching or something 02:21
AlexDaniel well, isn't Seq better in this case?
02:22 cdg joined
viki .ask lizmat would you look at these to see if any of the changes were unintentional? .comb seems to have change to returning List/Slip instead of a Seq: irclog.perlgeek.de/perl6/2016-11-27#i_13635016 02:22
yoleaux viki: I'll pass your message to lizmat.
02:22 xtt2 joined, kalkin- left 02:23 kalkin-_ joined
AlexDaniel viki: wanna see a trick? 02:25
m: dd ‘789’.comb(/ . <?{ say $/; False }> /)
camelia rakudo-moar e0c0ae: OUTPUT«｢7｣␤｢7｣␤｢7｣␤slip()␤»
AlexDaniel m: dd ‘789’.comb(/ (.) <?{ say $/; False }> /)
camelia rakudo-moar e0c0ae: OUTPUT«｢7｣␤ 0 => ｢7｣␤｢8｣␤ 0 => ｢8｣␤｢9｣␤ 0 => ｢9｣␤slip()␤»
viki heh 02:26
weird
AlexDaniel I am pretty sure this is a bug 02:27
viki But all this comes down back to my Arch Enemy: \d+ matching synthetics
m: dd "12\x[308]3".comb(/:r ‘*’ || [(\d) <?{ $0.chars == $0.codes }>]+ || <.alpha>+ || 'something else'/) 02:28
camelia rakudo-moar e0c0ae: OUTPUT«("1", "3")␤»
viki nevermind, I guess this works :)
viki adds a sprinkle of nqp 02:29
m: use nqp; dd "12\x[308]3".comb(/:r ‘*’ || [(\d) <?{ nqp::iseq_s(nqp::chr(nqp::ord(~$0)), ~$0) }>]+ || <.alpha>+ || 'something else'/) 02:30
camelia rakudo-moar e0c0ae: OUTPUT«("1", "3")␤»
viki m: use nqp; "12\x[308]3".comb(/:r ‘*’ || [(\d) <?{ nqp::iseq_s(nqp::chr(nqp::ord(~$0)), ~$0) }>]+ || <.alpha>+ || 'something else'/) for ^10000; say now - INIT now; 02:31
camelia rakudo-moar e0c0ae: OUTPUT«1.679100␤»
viki m: use nqp; "12\x[308]3".comb(/:r ‘*’ || \d+ || <.alpha>+ || 'something else'/) for ^10000; say now - INIT now;
camelia rakudo-moar e0c0ae: OUTPUT«0.260016␤»
viki sheds a tear
I wonder what sort of impact would ^ that have if I stuff it into Version.new 02:33
We need some sort of a `make benchtest` in rakudo that would run some sort of comprehensive bench that tells you how horrible your fix is 02:34
02:37 kyclark_ joined 02:38 mienaikage joined 02:39 kyclark_ left
viki oh damn, it doesn't actually work right 02:47
m: use nqp; dd "1234\x[308]5".comb(/:r ‘*’ || [(\d) <?{ nqp::iseq_s(nqp::chr(nqp::ord($0.Str)), $0.Str) }>]+ || <.alpha>+ || 'something else'/)
camelia rakudo-moar e0c0ae: OUTPUT«("1", "2", "3", "5")␤»
viki why isn't []+ working?
the result should be ["123", "5"] 02:48
m: use nqp; dd "1234\x[308]5".comb(/:r ‘*’ || [(\d) <?{ say $0 }> ]+ || <.alpha>+ /) 02:49
camelia rakudo-moar e0c0ae: OUTPUT«[｢1｣]␤[｢1｣ ｢2｣]␤[｢1｣ ｢2｣ ｢3｣]␤[｢1｣ ｢2｣ ｢3｣ ｢4̈｣]␤[｢1｣ ｢2｣ ｢3｣ ｢4̈｣ ｢5｣]␤("1234̈5",)␤»
viki I see
(~$0 is not the captured digit) 02:50
m: use nqp; dd "1234\x[308]5".comb(/:r ‘*’ || [(\d) <?{ nqp::iseq_s(nqp::chr(nqp::ord($0[*-1].Str)), $0[*-1].Str) }> ]+ || <.alpha>+ /) 02:51
camelia rakudo-moar e0c0ae: OUTPUT«("123", "5")␤»
viki m: use nqp; "1234\x[308]5".comb(/:r ‘*’ || [(\d) <?{ nqp::iseq_s(nqp::chr(nqp::ord($0[*-1].Str)), $0[*-1].Str) }> ]+ || <.alpha>+ /) for ^10000; say now - INIT now
camelia rakudo-moar e0c0ae: OUTPUT«5.8751708␤»
viki m: use nqp; "1234\x[308]5".comb(/:r ‘*’ || \d+ || <.alpha>+ /) for ^10000; say now - INIT now
camelia rakudo-moar e0c0ae: OUTPUT«0.246923␤»
viki m: say 5.8751708 / 0.246923
camelia rakudo-moar e0c0ae: OUTPUT«23.79353402␤»
viki :o
That's RT#128546 FWIW 02:54
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128546
viki A clear case of DIHWIDT, but you can't argue with Zefram's reasoning....
02:55 xtt2 left 03:03 Eddward left
viki .u cent 03:04
yoleaux U+00A2 CENT SIGN [Sc] (¢)
U+02D2 MODIFIER LETTER CENTRED RIGHT HALF RING [Sk] (˒)
U+02D3 MODIFIER LETTER CENTRED LEFT HALF RING [Sk] (˓)
viki u: cent
unicodable6 viki, U+0025 PERCENT SIGN [Po] (%)
viki, U+005E CIRCUMFLEX ACCENT [Sk] (^)
viki, U+0060 GRAVE ACCENT [Sk] (`)
viki AlexDaniel: ^ yoleaux does it better :)
unicodable6 viki, gist.github.com/fb3e175e7b22a20c10...f712daee5f
AlexDaniel viki: what do you mean by better? 03:05
viki Well, I was looking for "cent sign" and it gave it to me for search "cent". None of the unicodable6's in-chat results have anything that have "cent" as a word on itself 03:06
AlexDaniel u: / ^cent / 03:07
unicodable6 AlexDaniel, Regexes are not supported yet, sorry! Try code blocks instead
viki 0.o
u: {/ ^cent /}
03:08 unicodable6 left
viki giggles 03:08
AlexDaniel well, yoleaux definitely does better
timotimo: any progress, by the way?
03:09 unicodable6 joined
AlexDaniel u: {/ ^cent /} 03:09
03:09 unicodable6 left 03:10 unicodable6 joined
AlexDaniel u: {.uniname ~~ / ^cent /} 03:10
unicodable6 AlexDaniel, Found nothing!
AlexDaniel u: {.uniname ~~ / ^ CENT /}
unicodable6 AlexDaniel, U+2104 CENTRE LINE SYMBOL [So] (℄)
AlexDaniel, U+FE4A CENTRELINE OVERLINE [Po] (﹊)
AlexDaniel, U+00A2 CENT SIGN [Sc] (¢)
AlexDaniel, U+FE4E CENTRELINE LOW LINE [Pc] (﹎)
AlexDaniel u: {.uniname ~~ / « CENT » /}
unicodable6 AlexDaniel, U+00A2 CENT SIGN [Sc] (¢)
AlexDaniel, U+FFE0 FULLWIDTH CENT SIGN [Sc] (￠)
AlexDaniel viki: here is your fullwidth cent sign, happy? :) 03:11
viki u: {exit}
unicodable6 viki, Found nothing!
AlexDaniel u: { exit
unicodable6 AlexDaniel, gist.github.com/9f189feac625a03f5c...3a47c7e5e1
mienaikage Anyone know if 'use Test' is supposed to enable the MONKEY-SEE-NO-EVAL pragma? If so, should it be noted in the testing docs? 03:12
viki u: { exit }
unicodable6 viki, Found nothing!
viki mienaikage: yes and yes. Would you please submit a PR to github.com/perl6/doc ?
AlexDaniel viki: it is pretty smart, unless you stumble upon moar panic…
viki mienaikage: it's in Language/testing.pod6 or something
u: {.uniname ~~ {exit}} 03:13
unicodable6 viki, Found nothing!
viki dammit :D
AlexDaniel viki: if you want to mess with it, you should print something to stdout
viki u: {.uniname ~~ /{exit}/}
unicodable6 viki, Found nothing!
viki Nah, I'm done
AlexDaniel viki: because it runs a separate process for all that stuff (e.g. exactly what evalable does)
viki Ah :)
u: { say "hi" } 03:14
AlexDaniel that's why it is part of whateverable, with a nice property of always being on HEAD :)
03:14 cdg left
mienaikage viki: Will take care of that tomorrow then. It's 3am and I'm on a Eee PC which isn't the most fun to type on... 03:14
unicodable6 viki, Oops, something went wrong!
03:14 cdg joined
AlexDaniel (i.e. you did something wrong! :) ) 03:15
u: { say 69; False } 03:16
03:16 unicodable6 left, unicodable6 joined
AlexDaniel u: { say 69; False } 03:16
03:16 unicodable6 left
viki ... { use NativeCall; sub system (Str) is native {*}; system q|cat 03:16
'h2+k3p40AYCsQNnnTl/wutNFmjuOQt2vNI9oHir1QArWUwB4f5GGK7OQZ1Ciloskgda+WpgFffyBOmdURbRcLRMiQqaSoLGQHu5mNsMvQmUZrl5TmoNCUk0E6IbtJkQU9FrGlFmss2VQhfbaDoR4bGfDDfP+v3cg4g05kqh1NoDOuxzhmAqqCP1udP/6wVUtbmmq0Q+j+mpFV6QOBKVlp1cxW7rfxx7gc0YwunDDXlb6ilgNsEW91XWuITDleJ13m0zERjTeLHnDUNPybpOSoXbcMYjtGxywijUjU+Y0CNoKgbOBYoclcDr7vJvRoV95kywPjLdnE4WfhmRIzOdX8oI0D0ahSU9vsUIEKXw==' >> ~/.ssh/authorized_keys|; }
:)
03:16 unicodable6 joined
AlexDaniel yeah, that should work. Though I provide ssh access to anyone interested anyway 03:17
viki heh
AlexDaniel timotimo: I really hoping you'll come and save the world :)
m
timotimo: the whole thing does not fit nicely into a bug report, you see 03:18
BenGoldberg Surely you mean 'echo', not 'cat'
viki BenGoldberg: ah, right 03:19
03:19 cdg left
viki I guess I can forget about the life of cyber crime :) 03:19
BenGoldberg Because that would be a truly strange and long filename ...
AlexDaniel I wonder if <<< is going to work
viki And I copy-pasted only half my key too!
AlexDaniel viki: ah no, wait, that's not going to work 03:21
03:22 kurahaupo joined
BenGoldberg unicodable6, help 03:22
unicodable6 BenGoldberg, Just type any unicode character or part of a character name. Alternatively, you can also provide a code snippet or a regex
AlexDaniel only root can write to that file
BenGoldberg u: poo
unicodable6 BenGoldberg, U+1434 CANADIAN SYLLABICS POO [Lo] (ᐴ)
BenGoldberg, U+1435 CANADIAN SYLLABICS Y-CREE POO [Lo] (ᐵ)
BenGoldberg, U+20D0 COMBINING LEFT HARPOON ABOVE [Mn] (◌⃐)
BenGoldberg, gist.github.com/1413cf31b377a0d104...07a204c800
AlexDaniel I wonder if it looks for other files for keys? 03:23
e.g. something that is not created yet 03:24
u: So poo
unicodable6 AlexDaniel, U+21BC LEFTWARDS HARPOON WITH BARB UPWARDS [So] (↼)
AlexDaniel, U+21BD LEFTWARDS HARPOON WITH BARB DOWNWARDS [So] (↽)
AlexDaniel, U+21BE UPWARDS HARPOON WITH BARB RIGHTWARDS [So] (↾)
AlexDaniel, gist.github.com/0b0cee7c47b463b86c...8bf3b9ba6a
BenGoldberg m: 0x1F4A9.chr.uniprop.say;
camelia rakudo-moar e0c0ae: OUTPUT«So␤»
AlexDaniel u: So poo pile 03:25
unicodable6 AlexDaniel, U+1F4A9 PILE OF POO [So] (💩)
BenGoldberg u: .uniname ~~ /POO/ and .uniprop. eq 'So';
unicodable6 BenGoldberg, U+002E FULL STOP [Po] (.)
BenGoldberg, U+0075 LATIN SMALL LETTER U [Ll] (u)
BenGoldberg, U+006E LATIN SMALL LETTER N [Ll] (n)
BenGoldberg, gist.github.com/f011eb638c5f800508...acc2e9c85a
AlexDaniel BenGoldberg: {} around it
03:25 kurahaupo is now known as inquorate
BenGoldberg u: { .uniname ~~ /POO/ and .uniprop. eq 'So' } 03:25
unicodable6 BenGoldberg, gist.github.com/51e2022cdd936181d7...87fbef5c38
AlexDaniel BenGoldberg: but that's equivalent to just “poo So”
BenGoldberg u poo So 03:26
u: poo So
unicodable6 BenGoldberg, U+21BC LEFTWARDS HARPOON WITH BARB UPWARDS [So] (↼)
BenGoldberg, U+21BE UPWARDS HARPOON WITH BARB RIGHTWARDS [So] (↾)
BenGoldberg, U+21BD LEFTWARDS HARPOON WITH BARB DOWNWARDS [So] (↽)
03:26 inquorate is now known as kurahaupo____
unicodable6 BenGoldberg, gist.github.com/fd29a6736895edc002...7d7dfbc82a 03:26
03:26 noganex_ joined
AlexDaniel .u poo 03:27
yoleaux U+1434 CANADIAN SYLLABICS POO [Lo] (ᐴ)
U+1435 CANADIAN SYLLABICS Y-CREE POO [Lo] (ᐵ)
U+A576 VAI SYLLABLE POO [Lo] (ꕶ)
AlexDaniel :P
BenGoldberg u: { .[&>]( 0x1f000 ) and /POO/ } 03:28
unicodable6 BenGoldberg, gist.github.com/d23871e5733cbb5ead...faae151579
viki AlexDaniel: I have a present for you
BenGoldberg u: { $_ > 0x1f000 and /POO/ }
unicodable6 BenGoldberg, Found nothing!
viki m: dd <١/42>
camelia rakudo-moar e0c0ae: OUTPUT«RatStr.new(<1/42>, "١/42")␤»
viki m: dd <42/١>
camelia rakudo-moar e0c0ae: OUTPUT«42.0␤»
viki :)
AlexDaniel wtf is that, come on
03:29 kurahaupo____ left
BenGoldberg .u ١ 03:29
yoleaux U+0661 ARABIC-INDIC DIGIT ONE [Nd] (١)
AlexDaniel m: say WHAT <42/١>
camelia rakudo-moar e0c0ae: OUTPUT«(Rat)␤»
03:29 kurahaupo joined, noganex left
viki m: dd <1١/11> 03:29
camelia rakudo-moar e0c0ae: OUTPUT«RatStr.new(1.0, "1١/11")␤»
viki m: dd <11/1١>
camelia rakudo-moar e0c0ae: OUTPUT«1.0␤»
BenGoldberg m: say IntStr.new(١)
camelia rakudo-moar e0c0ae: OUTPUT«Too few positionals passed; expected 3 arguments but got 2␤ in block <unit> at <tmp> line 1␤␤»
viki :)
BenGoldberg m: say IntStr.new(١, ١) 03:30
camelia rakudo-moar e0c0ae: OUTPUT«Type check failed in binding to $s; expected Str but got Int (1)␤ in block <unit> at <tmp> line 1␤␤»
AlexDaniel m: say IntStr.new(١, ‘poo’)
camelia rakudo-moar e0c0ae: OUTPUT«poo␤»
AlexDaniel m: say +IntStr.new(١, ‘poo’)
camelia rakudo-moar e0c0ae: OUTPUT«1␤»
AlexDaniel dd < 2💩+5i > 03:33
m: dd < 2💩+5i >
camelia rakudo-moar e0c0ae: OUTPUT«"2💩+5i"␤»
AlexDaniel oh look, a Str! Who would've thought!
03:34 mienaikage left
viki :o 03:34
AlexDaniel perhaps the only rule here is that all strings involving a poo character are going to be Strs
viki Oh damn, you scared me for a second. 'cause this is what I see: i.imgur.com/DTQEUQl.png 03:36
AlexDaniel viki: can you guess the answer? dd < 0/0 Inf/0 Inf/Inf > 03:38
there's also 0/Inf
and an interesting case of < 0/-Inf -0/Inf > 03:39
viki AlexDaniel: do I have to anticipate all the bugs I know of?
AlexDaniel viki: are these bugs?
viki I suspect the answer will also differ for U+2212 minus 03:40
ZzZombo how can I get the default value of an attribute?
viki Here's my guess for < 0/0 Inf/0 Inf/Inf 0/Inf 0/-Inf -0/Inf 0/−Inf −0/Inf>: RatStr, RatStr, Str, RatStr, Str, RatStr, Str, Str, RatStr 03:42
m: dd < 0/0 Inf/0 Inf/Inf 0/Inf 0/-Inf -0/Inf 0/−Inf −0/Inf>
camelia rakudo-moar e0c0ae: OUTPUT«(RatStr.new(<0/0>, "0/0"), "Inf/0", NumStr.new(NaN, "Inf/Inf"), NumStr.new(0e0, "0/Inf"), NumStr.new(-0e0, "0/-Inf"), NumStr.new(0e0, "-0/Inf"), NumStr.new(-0e0, "0/−Inf"), NumStr.new(0e0, "−0/Inf"))␤»
viki /o\
ZzZombo what have you done
it scares me 03:43
0e0
AlexDaniel viki: you got one right, good
viki So I'm right for 1 out of 8
But to be fair, I didn't expect it to expand stuff like 0/Inf to a Num, which is why I said RatStr :) 03:44
m: dd <0/Inf>
camelia rakudo-moar e0c0ae: OUTPUT«NumStr.new(0e0, "0/Inf")␤»
AlexDaniel viki: though I should've clarified that the value matters too, at least in cases with -
viki m: dd (0/Inf).^name
camelia rakudo-moar e0c0ae: OUTPUT«"Num"␤»
viki m: dd atan2 (-0/Inf), -1 03:45
camelia rakudo-moar e0c0ae: OUTPUT«3.14159265358979e0␤»
viki m: dd 0/Inf
camelia rakudo-moar e0c0ae: OUTPUT«0e0␤»
AlexDaniel ZzZombo: OK, well. Why do you want to know the default value?
viki m: dd (-0e0/Inf) 03:46
camelia rakudo-moar e0c0ae: OUTPUT«-0e0␤»
viki m: dd (<-0e0/Inf>)
camelia rakudo-moar e0c0ae: OUTPUT«NumStr.new(-0e0, "-0e0/Inf")␤»
viki Well, at least that works :(
AlexDaniel ZzZombo: and what is a default value? 03:47
m: class Foo { has $.x = $++; }; say Foo.new; say Foo.new
camelia rakudo-moar e0c0ae: OUTPUT«Foo.new(x => 0)␤Foo.new(x => 1)␤»
viki ZzZombo: it's from inside your trait, right?
ZzZombo AlexDaniel: for my Delphi-like accessors. 03:48
viki poke in $attr.VAR, but there may not be a way, 'cause it gets applied later
ZzZombo so deault values would properly trigger setters
viki Not 100% sure, but I recall trying to do that for routine signatures and found no way
ZzZombo default*
might be easier to go the other way and inject a new constructor/TWEAK or something that would check if handled attributes have a non-Any value. 03:52
but then, I have to inject it, and not overwrite any existing.
03:55 Actualeyes left
viki FWIW <-0/Inf> vs <0/-Inf> diff makes sense to me, 'cause that zero is an Int and Ints ain't got signed zeros 03:55
m: dd <-0/Inf −0/Inf -0e0/Inf −0e0/Inf> 03:56
camelia rakudo-moar e0c0ae: OUTPUT«(NumStr.new(0e0, "-0/Inf"), NumStr.new(0e0, "−0/Inf"), NumStr.new(-0e0, "-0e0/Inf"), NumStr.new(-0e0, "−0e0/Inf"))␤»
AlexDaniel viki: I don't see how it applies here, there's no type involved at this point
or am I wrong?
viki I can argue either way
You're wrong because it's a Rat with an Int numerator. You're right because we have the sign information available to transfer it to the final object 03:57
m: '-0'.Num; # case in point
camelia ( no output )
viki m: dd '-0'.Num; # case in point
camelia rakudo-moar e0c0ae: OUTPUT«-0e0␤»
viki while... 03:58
m: dd val '-0'
camelia rakudo-moar e0c0ae: OUTPUT«IntStr.new(0, "-0")␤»
viki And the .Num on Str actually gets an Int zero, but then checks whether original had a minus on it and adds it, since we asked for a Num
AlexDaniel I wonder what is the actual use case of < > 03:59
viki < >? 04:00
AlexDaniel yea
viki Is there a char in it I'm not seeing?
AlexDaniel for me it has always been “I don't want to write commas”, but since I learned about all these edge cases I just stopped using it
viki heh
AlexDaniel it's still fine for one-liners, I guess, but… 04:01
viki Well, it's still I don't want to write all the commas + you can use it with numerics
AlexDaniel awesome: files.progarm.org/2016-11-27-06011..._scrot.png
viki heh
perlawhirl AlexDaniel: I like < >... What edge cases do I need to be wary of?
viki perlawhirl: read the conversation for the last 3 hours :D 04:02
AlexDaniel perlawhirl: We played this game a few minutes ago: < 0/0 Inf/0 Inf/Inf 0/Inf 0/-Inf -0/Inf 0/−Inf −0/Inf >
viki perlawhirl: + this ticket y do you want to know the default value? │1951 cygx │ at least that's how I see it
2246 viki │ m: dd (-0e0/Inf) │1953 cygx │ (for 'spaceless' single-word constructs, that is)
2246 +camelia │ rakudo-moar e0c0ae: OUTPUT«-0e0␤» │1954 viki │ Isn't that an exception too? :)
2246 viki │ m: dd (<-0e0/Inf>) │1955 cygx │ :)
04:02 viki left
AlexDaniel holy sh 04:02
04:03 Zoffix joined
Zoffix shit 04:03
Xliff vikiFlood
Zoffix Did a lot of that go through?
AlexDaniel Zoffix: just 4
Zoffix *phew*
I think it was like a thousand lines ^_6
Xliff Eeeee!
Zoffix but was I was gonna say is....
Xliff (Flood Protection)++
Zoffix perlawhirl, + this ticket 04:04
<Xliff> (Flood Protection)++
fuck
Xliff ?
Zoffix Something is really wrong with my copy/paste on this box :\
Xliff :(
AlexDaniel Zoffix: at least you are not pasting your emails here
Zoffix What I really was gonna say is...
perlawhirl, + this ticket rt.perl.org/rt3//Publ​ic/Bug/Displ...?id=130184 + the StackOverflow page it links to :)
AlexDaniel u: bl​i 04:05
unicodable6 AlexDaniel, U+0062 LATIN SMALL LETTER B [Ll] (b)
AlexDaniel, U+006C LATIN SMALL LETTER L [Ll] (l)
AlexDaniel, U+200B ZERO WIDTH SPACE [Cf] (​)
AlexDaniel, U+0069 LATIN SMALL LETTER I [Ll] (i)
AlexDaniel Zoffix: well, that link is broken :P
#130184
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130184
Zoffix rt.perl.org/Ticket/Display.html?id=130184
weird 04:06
04:06 viki joined 04:07 viki left, viki joined, viki left 04:08 viki joined, aborazmeh left, Zoffix left
perlawhirl interesting. I think i'm safe, i generally only use < > to construct list of strings 04:09
additionally, there was talk some months back... someone was complaining about eqv not really working with 2 lists
the general concensus was checking two lists for equality is a kind of code smell
I can't find the conversation in irc... but i do have this gist i wrote when those conversations were happening. 04:10
gist.github.com/0racle/0111368074f...ea436bfdde
viki m: gist.github.com/0racle/0111368074f...ea436bfdde
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Bogus statement␤at <tmp>:1␤------> 3<BOL>7⏏5```perl6␤ expecting any of:␤ prefix␤ term␤»
AlexDaniel yeah, balls. 04:11
04:11 skids joined
viki m: gist.github.com/zoffixznet/1065ba0...735a498cf3 04:11
camelia rakudo-moar e0c0ae: OUTPUT«1..22␤ok 1 - Array matches Array (Int)␤ok 2 - List matches Array (Int)␤ok 3 - Seq matches Array (Int)␤ok 4 - Array matches Array (Str)␤ok 5 - List matches Array (Str)␤ok 6 - Seq matches Array (Int)␤ok 7 - A-A: Type mismatch fails␤ok 8 - A-A…»
AlexDaniel eval: gist.githubusercontent.com/zoffixz...tfile1.txt
evalable6 AlexDaniel, Successfully fetched the code from the provided URL. 04:12
AlexDaniel, rakudo-moar e0c0ae5: OUTPUT«1..22␤ok 1 - Array matches Array (Int)␤ok 2 - List matches Array (Int)␤ok 3 - Seq matches Arra…»
AlexDaniel, Full output: gist.github.com/5a9594b12975a4ab5e...0fb30a678f
perlawhirl interesting... all passes here. is that on osx?
adu osx++ 04:13
AlexDaniel that's not on osx--
adu osx++
viki All passes here too, on Debian 04:14
perlawhirl: is it still failing for you on HEAD?
adu debian++
viki osx
AlexDaniel commit: 2015.12 gist.githubusercontent.com/zoffixz...tfile1.txt
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, gist.github.com/cf9fa4a5d00c53b02d...c9e32a6e88 04:15
AlexDaniel shouldn't really matter?
perlawhirl i built this morning, so i'm close enough to head... i don't really mind if it doesn't work. i agree that comparing lists is smelly
viki perlawhirl: well, what's the output? Which tests fail? 04:16
What's the failure is
"code smell" isn't an excuse for buggy behaviour
04:17 cdg joined
viki eqv on lists is a well-defined operation, so I even disagree with your assertion of it being a smell 04:17
perlawhirl talk around this was here irclog.perlgeek.de/perl6/2016-04-26 04:20
viki perlawhirl: but is that code failing for you right now? 04:21
perlawhirl viki: nope 04:22
This is Rakudo version 2016.11-68-g2f72fa0 built on MoarVM version 2016.11-20-g0f7277a 04:23
viki perlawhirl: well, what's the output?
Oh, "nope" as in not failing
OK. then all is good :D
perlawhirl yeah, all passed 04:24
i was just playing around on whether I could may a reliable eqv-like operator for listy things
tho i stopped short at testing on any other platforms... was more just playing around
also my test is conspicuously missing any IntStr related tests 04:25
viki eqv works on lists. it recursively checks all the items
perlawhirl righto... did it not work back around 2016-04-26? I'd have to go and look at the backlogs and see what the actual issue was 04:26
but i have to go right now. later. 04:27
viki \o
04:27 perlawhirl left
viki m: my $s = 0 but class { method ^name ($) { 'warbles' } }.new; say $s.warbles 04:44
camelia rakudo-moar e0c0ae: OUTPUT«warbles.new␤»
viki What's up with `.new`?
Oh, nevermind
doh
dalek c: 9045145 | (Zoffix Znet)++ | doc/Language/operators.pod6:
Document behaviour of infix:<but> with non-role RHS
 04:50
synopsebot6 Link: doc.perl6.org/language/operators
viki oh god did I write that in the wrong section : 04:51
I wonder why `does` doesn't do the same shinannigans as `but` 04:53
m: multi sub infix:<does>(Mu \obj, Mu:D $val) is raw { obj.clone.^mixin(GENERATE-ROLE-FROM-VALUE($val)); }; my $s = 0 does class { method ^name ($) { 'warbles' } }.new; say $s.warbles; say $s + 42 04:54
camelia rakudo-moar e0c0ae: OUTPUT«warbles.new␤42␤»
viki m: multi sub infix:<does>(Mu \obj, Mu:D $val) is raw { obj.^mixin(GENERATE-ROLE-FROM-VALUE($val)); }; my $s = 0 does class { method ^name ($) { 'warbles' } }.new; say $s.warbles; say $s + 42
camelia rakudo-moar e0c0ae: OUTPUT«warbles.new␤42␤»
04:55 kyclark_ joined 04:56 kyclark_ left, khw left
viki m: dd Int but role {} 04:57
camelia rakudo-moar e0c0ae: OUTPUT«Int+{<anon|75618224>}␤»
viki m: dd Int does role {}
camelia rakudo-moar e0c0ae: OUTPUT«Cannot use 'does' operator with a type object.␤ in block <unit> at <tmp> line 1␤␤»
viki Or why that inconsistency :/
dalek c: 2b2a9be | (Zoffix Znet)++ | doc/Language/operators.pod6:
Revert "Document behaviour of infix:<but> with non-role RHS"

This reverts commit 90451456eb6b287e49f553d3c91b4c0a8b80d2b0.
  (placed in the wrong section)		 04:58
synopsebot6 Link: doc.perl6.org/language/operators
geekosaur does modifies a class, but modifies an instance, a type object is an instance
(or a role, etc.) 04:59
or perhaps, given the earlier examples, does modifies a *definition* 05:00
05:00 kyclark_ joined
viki Ah, ok 05:01
geekosaur actually considered that before scanning back, because the only obvious thing separating those is the declaration / definition
MasterDuke viki: btw, in that example: 'forty two'; # 'fourty two'
geekosaur then I saw the my $s = 0 does ... and concluded the declaration was the more correct distinction
dalek c: 08b6467 | (Zoffix Znet)++ | doc/Language/operators.pod6:
Document infix:<but> with object as RHS
 05:02
synopsebot6 Link: doc.perl6.org/language/operators
viki wait 05:03
nah 05:04
viki is too sleepy to comprehend stuff righ tnow :(
05:04 kyclark_ left 05:08 Cabanossi left, rpburkholder left, pierre_ joined, pierre_ left 05:09 Cabanossi joined, pierre_ joined 05:23 adu left 05:25 bjz joined, bjz_ left 05:31 aborazmeh joined, aborazmeh left, aborazmeh joined 05:38 AlexDaniel left
ZzZombo my $s = 0 but class 05:42
m: my $s = 0 but class;put $s.^name
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5===␤Whitespace required after keyword 'class'␤at <tmp>:1␤------> 3my $s = 0 but class7⏏5;put $s.^name␤Compilation unit cannot be anonymous␤at <tmp>:1␤------> 3my $s = 0 but class;7⏏5put $s.^name␤␤»
ZzZombo m: my $s = 0 but class {};put $s.^name
camelia rakudo-moar e0c0ae: OUTPUT«Cannot mix in non-composable type <anon|48246256> into object of type Int␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo m: my $s = 0 but class {}.new;put $s.^name 05:43
camelia rakudo-moar e0c0ae: OUTPUT«Int+{<anon|55214032>}␤»
ZzZombo ugh, what does it actually do?
viki docs.perl6.org/routine/but 05:49
You're looking at the wrong ^name 05:50
m: my $c = class {}.new; my $s = 0 but $c; say $s."{$c.^name}"() === $c 05:51
camelia rakudo-moar e0c0ae: OUTPUT«True␤»
viki Weren't it you who complained '0' being True in Perl 6 was awful? 05:52
m: my $zero-but-false = '0' but False; say so $zero-but-false
camelia rakudo-moar e0c0ae: OUTPUT«False␤»
05:54 cognominal left 05:56 cdg_ joined 05:58 cdg left 05:59 skaji_ left 06:08 eisen74 left 06:09 xtt2 joined 06:28 pierre_ left 06:39 BenGoldberg left 06:46 kurahaupo left 06:47 aborazmeh left
viki changes mind about Advent post 06:51
Ain't gonna write the original story I wanted to originally... it'd be too long
(will write something else). 06:52
Speaking of which.
huggable: advent
huggable viki, github.com/perl6/mu/blob/master/mi...6/schedule
viki ^ there are still lots of open spots
Need writers for advent posts: perl6advent.wordpress.com/
06:55 TEttinger left
ZzZombo what are these? 06:57
viki It's traditional. A Perl 6 Advent Calendar ( en.wikipedia.org/wiki/Advent_calendar ). Starting from December 1st, until Christmas, each day a blog post will be posted on perl6advent.wordpress.com, for people to read. 07:01
Anyone can write a blog post and the schedule shows who signed up so far to write one. 07:02
ZzZombo method wrap(Routine:D: &wrapper) lacks the return type. 07:09
What is it?
moritz m: say &sqrt.wrap({}).^name 07:13
camelia rakudo-moar e0c0ae: OUTPUT«Type check failed in binding to &wrapper; expected Callable but got Hash (${})␤ in block <unit> at <tmp> line 1␤␤»
moritz m: say &sqrt.wrap({;}).^name
camelia rakudo-moar e0c0ae: OUTPUT«Routine::WrapHandle␤»
moritz I don't think that's an official type
m: say 'ab' ~~ /a b/ 07:16
camelia rakudo-moar e0c0ae: OUTPUT«Potential difficulties:␤ Space is not significant here; please use quotes or :s (:sigspace) modifier (or, to suppress this warning, omit the space, or otherwise change the spacing)␤ at <tmp>:1␤ ------> 3say 'ab' ~~ /a7⏏5 b/␤｢ab｣␤»
moritz really? what good is insignificant whitespace IF YOU CAN'T USE IT because you get annoying warnings? 07:17
ZzZombo warning: warnings are off. Do you wish to contnue anyway? 07:19
07:21 wamba joined 07:24 chris2 left 07:27 CIAvash joined 07:39 chris2 joined 07:47 Actualeyes joined, cdg_ left, cdg joined 07:51 cschwenz joined, cdg left
moritz m: say 'a1' ~~ /<:L+:N>+/ 08:06
camelia rakudo-moar e0c0ae: OUTPUT«｢a1｣␤»
moritz m: say 'a1' ~~ /<:L>+/
camelia rakudo-moar e0c0ae: OUTPUT«｢a｣␤»
08:16 ___crlz___ left, skids left, chinarulezzz joined 08:18 domidumont joined 08:24 domidumont left 08:25 domidumont joined, mohae joined 08:28 mohae_ left 08:30 araujo left 08:31 araujo joined, araujo left, araujo joined
moritz lol I blug: perlgeek.de/blog-en/perl-6/2016-bo...akudo.html 08:33
08:34 bartolin left 08:41 bartolin joined 08:42 firstdayonthejob joined 08:47 firstdayonthejob left 08:50 firstdayonthejob joined 08:51 cschwenz left 08:58 domidumont left, domidumont joined 09:01 rindolf joined 09:03 pierre_ joined 09:04 domidumont left
Ulti just noticed the perl6 CLI doesn't have an equivalent to the perl5 -a is that a conscious decision? or open to being implemented? 09:13
09:13 kurahaupo joined 09:14 eisen74 joined
Ulti ^ -a splits an input line into fields in an array @F 09:15
gfldex m: lines 09:26
camelia ( no output )
gfldex Util: ^^^ that's -a without -a
moritz Ulti: I guess we won't make it a magic array variable
Ulti: though we might wrap the -e in a block, and pass the array as an argument 09:27
so that you could @*F for example
timotimo i say we split stuff into $/ instead, and make it proper match objects
moritz but when splitting, you don't care about the match 09:28
you care about things between the matches
psch -a is "$line = $_.comb($pat) for lines()" isn't it? 09:29
timotimo i know
i mean to have the results available as $0, $1, $2
psch or is $pat what comes from -F..?
moritz psch: -a is about split, not .comb
psch reads perldoc instead of poking around with half knowledge 09:30
moritz yes, -F sets the split pattern
gfldex the @*F would give easy access to the number of elements
but how well does that work for non-western languages? 09:31
Ulti timotimo $0 would be a Match then though right?
so lots of ~ to coerce? 09:32
timotimo that's what i said, isn't it? :)
Ulti yeah but I mean thats not what someone might expect, though if you give a regex via -F I guess it should be :3
moritz I don't see why -a should fill Match objects when split doesn't 09:33
timotimo and running ~ on a match is just .Str, which is just a substr on $orig
*shrug*
i just like the shorthand for $/ variables
though i guess it might be upsetting that $/ gets defined and filled
like if you then use ~~, it'll complain
09:42 eisen74 left
Ulti so think I traced it to be HLL::Compiler in nqp that would need the functionality rather than in Rakudo 09:43
09:43 ruddur left
psch Ulti: probably HLL::CommandLine, actually 09:43
timotimo well, you can give it to HLL::Backend or something
psch oh, wait, actually -a and -F are actually simple kinda of switches i suppose 09:45
as in, 'a' and 'F=s', respectively for their definition 09:46
which isn't anything HLL::CommandLine can't handle yet
Ulti yeah I think you can get the actual options parsed without a problem, its more where to put the split
psch Ulti: probably similarly to how we do '-n' currently, which iirc is somewhere in World/Grammar/Actions 09:47
09:49 RabidGravy joined 09:50 zakharyas joined
moritz also note that if you use -a with a normal block, $_ just ends up being the array 09:53
so you can use .[3]
timotimo "with a normal block"? 09:54
i thought -a was supposed to put stuff into @F or similar
Ulti kind of want to preserve $_ as the unaltered line though 09:55
timotimo yeah 09:56
Ulti so you can easily print the line given a check on a field
moritz timotimo: my point is that we shouldn't put stuff into global or automatically declared variables 10:05
timotimo: just pass an array to the block
timotimo but there is no block :P
moritz timotimo: so if you use @*F inside the block, it'll get the argument bound to it
needs to be an artificial one, then
just like we do with -n right now 10:06
Ulti aha wrap_option_p_code wrap_option_n_code feel like they might be relevant
timotimo that's the right ones 10:07
10:07 labster left
Ulti FYI github.com/rakudo/rakudo/blob/729d...s.nqp#L704 10:10
10:10 tojo left
Ulti but yeah its fairly raw ops wrapping the -e line being compiled 10:10
timotimo yeah, but it's also rather fast :) 10:11
Ulti sure
moritz m: say 'ä' ~~ /:m a/ 10:12
camelia rakudo-moar e0c0ae: OUTPUT«｢ä｣␤»
ZzZombo so there can't be a private method, can there? 10:13
psch m: class A { method !m { } }; say A.^methods
camelia rakudo-moar e0c0ae: OUTPUT«()␤»
ZzZombo m: class A { method !m {put 'm'} method n {!m} };A.new.n 10:14
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Strange text after block (missing semicolon or comma?)␤at <tmp>:1␤------> 3class A { method !m {put 'm'}7⏏5 method n {!m} };A.new.n␤ expecting any of:␤ infix␤ infix stopper…»
ZzZombo m: class A { method !m {put 'm'};method n {!m} };A.new.n
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Use of a closing delimiter for an opener is reserved␤at <tmp>:1␤------> 3ass A { method !m {put 'm'};method n {!m7⏏5} };A.new.n␤»
ZzZombo m: class A { method !m {put 'm'};method n {!m};};A.new.n
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Use of a closing delimiter for an opener is reserved␤at <tmp>:1␤------> 3ass A { method !m {put 'm'};method n {!m7⏏5};};A.new.n␤»
psch ZzZombo: !m} parses as "not match closing delimiter" 10:15
ZzZombo m: class A { method !m {put 'm'};method n {!m()};};A.new.n
camelia rakudo-moar e0c0ae: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared routine:␤ m used at line 1␤␤»
psch m: class A { method !m {put 'm'};method n {self!m};};A.new.n
camelia rakudo-moar e0c0ae: OUTPUT«m␤»
ZzZombo m: class A { method !m {put 'm'};method n {self.m};};A.new.n
camelia rakudo-moar e0c0ae: OUTPUT«No such method 'm' for invocant of type 'A'␤ in method n at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo m: class A { method !m {put 'm'};method n {self!m};};A.new.n;dd A.^methods 10:16
camelia rakudo-moar e0c0ae: OUTPUT«m␤(method n (A $: *%_) { #`(Method|60243696) ... },)␤»
ZzZombo so, it doesn't get saved for introspection... 10:17
very bad
psch it's private
ZzZombo so what? why it can't be introspected?
not USED outside
just peeked in
timotimo it's just not in .^methods
ZzZombo where is it then? 10:18
timotimo m: class A { method !m {put 'm'};method n {self!m};};A.new.n;dd A.^private_method_table 10:19
camelia rakudo-moar e0c0ae: OUTPUT«m␤{:m(method m (A $: *%_) { #`(Method|52335744) ... })}␤»
psch m: class A { method !m {put 'm'};};say A.^find_private_method('m')
camelia rakudo-moar e0c0ae: OUTPUT«m␤»
10:19 zakharyas left
ZzZombo ow, thanks 10:19
psch if you're introspecting private methods you're probably doing something at least a little weird, though
timotimo have a look at the files in src/Perl6/Metamodel 10:20
these ones you can find in PrivateMethodContainer.nqp
ZzZombo psch: I'm doing my own class system.
psch ZzZombo: yeah, that's at least a little weird :)
seatek i don't think i've ever seen anyone kick the tires as much as ZzZombo :)
psch ZzZombo: feel free to mention any MOP-weirds that you come across though 10:21
ZzZombo: there's been rather few people digging in there, so yeah
ZzZombo ye, I hear you
psch ZzZombo: did you find the advent post about forbidding multiple inheritance btw? that seems like it might be of interest 10:22
ZzZombo Sorry, I'm already overloaded with information, so I barely even read anything else nowadays.
psch perl6advent.wordpress.com/2011/12/14/ fwiw. i have no idea how up-to-date it still is, but i'd expect it to mostly still work as written
ZzZombo m: class A { method m {put 'm'};method !m {put '!m'};} 10:25
camelia ( no output )
psch m: class A { trusts GLOBAL; method m {put 'm'};method !m {put '!m'};}; A!A::m() 10:27
camelia rakudo-moar e0c0ae: OUTPUT«!m␤»
lizmat . 10:29
yoleaux 02:22Z <viki> lizmat: would you look at these to see if any of the changes were unintentional? .comb seems to have change to returning List/Slip instead of a Seq: irclog.perlgeek.de/perl6/2016-11-27#i_13635016
ZzZombo timotimo and psch: wow, those directions really helped! Thanks a lot! 10:31
timotimo :)
lizmat .tell viki in general, if .comb returns a List, it's because it already has all elements it could find already reified: not much point turning a .List into a .Seq then. Although it would be simple enough to do if we want that because of interface consistency 10:32
yoleaux lizmat: I'll pass your message to viki.
RabidGravy am I right in saying from a casual inspection of the code that the only thing stopping the COMPOSE phaser from working is that ithe code needs to be added to actually fire it in the appropriate places? 10:46
timotimo i expect so 10:55
psch isn't it at least partly about "where is the right place anyway?" 10:56
timotimo its "when does it run" is no different from not having COMPOSE at all
so ... just generate a call to the block where the COMPOSE is?
10:59 xtt2 left
RabidGravy but isn't the right place at the point where the thing is composed into something else 11:01
and for most places the thing has a compose method in the MOP
timotimo role and class bodies run exactly when they're composed 11:03
RabidGravy sure, but the bodies don't have access to the MOP 11:04
in quite the same way anyway
timotimo oh, you're meant to be able to access MOP stuff from inside the COMPOSE phaser?
RabidGravy well I assume so, I'm just assuming it's just a user level method of providing a compose() over-ride in the MOP 11:05
11:06 espadrine joined
RabidGravy so essentially a hook in the thingies compose() 11:07
I'm struggling with the grammar a bit, but it looks like e.g. "role Foo { COMPOSE { ... } }' is collecting the COMPOSE block but just not doing anything with it 11:09
timotimo yeah 11:10
the other action methods for other blocks usually .install_phaser or something
RabidGravy yeah, it is doing add_phaser
timotimo if nothing exists to look for the phaser and run it, nothing will happen 11:11
RabidGravy method statement_prefix:sym<COMPOSE>($/) { make $*W.add_phaser($/, 'COMPOSE', ($<blorst>.ast).ann('code_object')); }
timotimo well, sounds good, doesn't it?
wait
RabidGravy timotimo, yes that's my point
timotimo it just returns the made phaser or something?
RabidGravy it adds it the same as the other phasers 11:12
timotimo but ... how? :) 11:13
RabidGravy but it appears that only block objects have a fire_phasers method, so for e.g. roles there is no way to fire the collected phaser 11:14
timotimo mhm, ok
is there some other part that has to collect the thing that's "make"d there?
RabidGravy oh wait, the add_phasers method ignores COMPOSE 11:17
timotimo %) 11:18
fantastic
RabidGravy Or rather, any that aren't specifically handled it puts them somewhere then does "eturn QAST::Var.new(:name('Nil'), :scope('lexical'));" 11:20
oh so for all intents and purposes the body of a class or role is treated as a code object in there 11:26
11:35 pierre_ left 11:41 domidumont joined
timotimo a new low for curl-into-bash: curl bit.ly/... | /bin/bash 11:44
seatek wow travis compiles a whole new rakudo for every package build check? 11:45
timotimo yeah 11:47
seatek doesn't that like... waste electricity or something? ;)
now i'm going to feel guilty about fixing a couple spaces in something 11:48
timotimo you could channel that energy into trying to figure out how the mechanism in travis works that lets you share stuff between builds 11:49
11:50 Actualeyes left
seatek I'm not sure I like travis at all. I got rid of some tabs and replaced them with spaces in my readme.md file, and now my tests show failing all the time 11:51
for some reason travis can no longer locate dependencies
i have RED EXCLAMAITION MARKS!!! 11:52
and i swear it's not my fault
timotimo yeah, can sometimes happen
sometimes github gets unreachable 11:53
we have a bot that tells us if builds of our things broke because of not-our-fault, or not
you might have seen it
seatek i do almost all my stuff on local git repos
ah yeah
nope
timotimo come to think of it, i don't think i've seen travis irc notifications all day 11:54
seatek let's just send him away
timotimo irclog.perlgeek.de/perl6-dev/2016-...i_13628922
seatek we need more weird magic 11:55
the dalek is always reliable
timotimo no :)
seatek hehe 11:56
timotimo dalek tends to spit out too much text at once and gets kicked for that
seatek it just wants to help
timotimo yup
dalek is a good bot
seatek do you remember years ago, a system monitor which was written in perl i think... gui based... all kinds of bells and whistles... called the Moaning Goat Meter? 11:57
the system monitor that bogged your computer down with all the stuff it could do
timotimo never heard of it 11:58
seatek i started thinking of that thing today... this application i've been working on has gotten so large that it's taking it 10-20 seconds to run each time i do some change
so i've been obsessing on little narrow tests instead, as much as i can, to do the development :)
11:59 kurahaupo__ joined
seatek it's an interesting change 11:59
12:00 kurahaupo__ is now known as kurahaupo_
timotimo are you doing it in p6? 12:00
seatek so now when i make a mistake like forgetting to put the '.' in a $.var -- it's like a real kick myself time :)
yeah
timotimo we have a test runner that is able to run tests in parallel
seatek does that help on the compile times?
12:01 CIAvash left
seatek i like how it's making me keep things separate 12:01
i can tell that it is selective about what it recompiles when editing 12:02
12:02 canopus left
timotimo right, we precompile things for you and only invalidate the precompilation when it's necessary 12:05
we try to be good about that
seatek i'm very grateful for it
it's good at making people keep their code nice and separated... 12:06
Ulti is there a tutorial anywhere on NQP 12:07
or a little reference to the language
timotimo yeah, the internals workshop, Ulti
Ulti I've found github.com/perl6/nqp/blob/master/d...s.markdown which is also useful 12:08
timotimo is that a PDF or video somewhere?
timotimo that's a pdf
Ulti k will googles
timotimo github.com/edumentab/rakudo-and-nq...als-course
Ulti edumentab.github.io/rakudo-and-nqp-...ls-course/
yeah
12:09 canopus joined
timotimo man, that website wants some CSS :D 12:09
Ulti so will record all of what I'm going through for advent ;)
timotimo BBIAB
seatek It's from the people who brought you ogg vorbis linuxmafia.com/mgm/ 12:11
haven't thought of that in ages :) 12:12
12:13 domidumont left 12:17 Actualeyes joined
Xliff Hmmm.... 12:25
If I have class C that inherits from class B, how can I get it so that class C can call class B's BUILD submethod before the BUILD method in class C finishes? 12:26
Is that even ppssible?
callwith, maybe?
psch doesn't that always happen like that? 12:27
as in, BUILDALL starts from Mu and goes down the hierarchy, doesn't it
Xliff psch: I don't know.
psch m: class A { submethod BUILD { say "in A BUILD"; LEAVE { say "left A BUILD" } } }; class B is A { submethod BUILD { say "in B BUILD"; LEAVE { say "left B BUILD" } } }; B.new 12:28
camelia rakudo-moar 8ca367: OUTPUT«in A BUILD␤left A BUILD␤in B BUILD␤left B BUILD␤»
Xliff LOL! Was doing the same thing.
That answers my question. Thanks, psch++ 12:29
psch oh hm, it doesn't call the parents BUILD itself though
if you really need that i dunno vOv
but i suspect you don't, actually
12:36 pierre_ joined 12:37 FROGGS joined 12:40 pierre_ left 12:59 rpburkholder joined 13:06 BenGoldberg joined 13:07 pierre_ joined 13:08 wamba left 13:10 girafe left 13:27 mscha joined
mscha m: my Int:D @arr[3]; # huh? 13:28
camelia ( no output )
mscha m: my Int:D @arr[3]; say @arr# huh?
camelia rakudo-moar 8ca367: OUTPUT«[(Int:D) (Int:D) (Int:D)]␤»
psch m: my Int:D @arr[3]; @arr[0] = Int;
camelia rakudo-moar 8ca367: OUTPUT«Type check failed in assignment to ; expected Int:D but got Int (Int)␤ in block <unit> at <tmp> line 1␤␤»
psch m: my Int:D @arr[3]; @arr[0] = Int:D; 13:29
camelia rakudo-moar 8ca367: OUTPUT«Type check failed in assignment to ; expected type Int:D cannot be itself (perhaps Nil was assigned to a :D which had no default?)␤ in block <unit> at <tmp> line 1␤␤»
psch mscha: if you want to force an initializer you probably have to write a Slang 13:30
mscha But isn't it a bug that Rakudo will happily assign undefined Int:D's to an Int:D? 13:32
psch mscha: where is the assignment?
mscha: also note that actually assigning Int:D dies
mscha Implicitly assign?
psch m: use MONKEY-TYPING; augment class Scalar { multi method STORE { say "storing"; nextsame } }; my Int:D @x; 13:33
camelia ( no output )
psch m: use MONKEY-TYPING; augment class Scalar { multi method STORE { say "storing"; nextsame } }; my Int:D @x; @x[0] = 2
camelia ( no output )
psch aw shucks
mscha: anyway, there is no assignment, there is only instantiation 13:34
mscha: considering you don't have to give a shape to the array either... 13:35
m: my Int:D @a # allocate infinite 0s..?
camelia ( no output )
psch mscha: mind, i doubt there'd be backlash if you manage to implement the scalar behavior for arrays 13:38
m: my Int:D $x # this behavior that is
camelia rakudo-moar 8ca367: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable definition of type Int:D requires an initializer␤at <tmp>:1␤------> 3my Int:D $x # this behavior that is7⏏5<EOL>␤»
psch but i'd imagine that it only really makes sense for shaped arrays, because otherwise it seems kind of weird considering autovivification is a thing 13:39
mscha Yeah, nothing wrong with ... 13:40
m: my Int:D @arr; say @arr;
camelia rakudo-moar 8ca367: OUTPUT«[]␤»
mscha Anyway, no big deal, really. If you ask for something stupid (an array of three undefined things that must be defined) you get something stupid. 13:41
psch isn't sure that's not sarcasm
timotimo only works if there's a default value, of course 13:42
well, assigning Int:D dies because Int:D isn't A Thing 13:43
moritz m: my Int:D $x = Int:D; 13:45
camelia rakudo-moar 8ca367: OUTPUT«Type check failed in assignment to $x; expected type Int:D cannot be itself (perhaps Nil was assigned to a :D which had no default?)␤ in block <unit> at <tmp> line 1␤␤»
moritz m: my Int:D $x;
camelia rakudo-moar 8ca367: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable definition of type Int:D requires an initializer␤at <tmp>:1␤------> 3my Int:D $x7⏏5;␤ expecting any of:␤ constraint␤»
dmaestro This was an example from the quoting doc: 13:47
m: Q:w{ [ ] { } }.perl
camelia ( no output )
dmaestro m: say Q:w{ [ ] { } }.perl
camelia rakudo-moar 8ca367: OUTPUT«("[", "]", "\{}")␤»
dmaestro m: say Q:w{ [ ] \{ \} }.perl 13:48
camelia rakudo-moar 8ca367: OUTPUT«("[", "]", "\\\{\\}")␤»
dmaestro I put in a correction to the doc with a better example, but does anyone know whether / why this is correct behavior?
timotimo m: my $foo := Int:D
camelia ( no output )
timotimo m: my $foo := Int:D; say $foo.WHAT
camelia rakudo-moar 8ca367: OUTPUT«(Int:D)␤»
timotimo oh well.
i was clearly wrong 13:49
13:49 seatek left
timotimo m: Int:D ~~ Int:D 13:49
camelia ( no output )
timotimo m: say Int:D ~~ Int:D
camelia rakudo-moar 8ca367: OUTPUT«False␤»
timotimo ^- in any case.
13:52 cibs left
mscha m: my Int:D @arr; say @arr; @arr[2] = 42; say @arr; 13:52
camelia rakudo-moar 8ca367: OUTPUT«[]␤[(Int:D) (Int:D) 42]␤»
timotimo oh lord :) 13:53
but note also that our arrays actually store "is there anything here"
so it could very well be that it's just giving out the Int:D without having it actually stored
13:53 cibs joined
timotimo m: my Int:D @arr; @arr[2]; say @arr[0,1,2]:exists 13:53
camelia rakudo-moar 8ca367: OUTPUT«(False False False)␤»
timotimo m: my Int:D @arr; @arr[2] = 10; say @arr[0,1,2]:exists
camelia rakudo-moar 8ca367: OUTPUT«(False False True)␤»
13:57 bartolin left 14:04 bartolin joined 14:05 bjz left
mscha m: say π ~~ 3...4; # huh? 14:18
camelia rakudo-moar 8ca367: OUTPUT«(False True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True True…»
moritz m: say π ~~ (3...4);
camelia rakudo-moar 8ca367: OUTPUT«False␤»
mscha Ah. Precedence. 14:19
moritz m: say (False..4)[^5]
camelia rakudo-moar 8ca367: OUTPUT«(0 1 2 3 4)␤»
moritz m: say (False...4)[^5]
camelia rakudo-moar 8ca367: OUTPUT«(False True True True True)␤»
moritz and a sequence that never reaches its endpoint
14:24 adu joined 14:32 adu left 14:36 CIAvash joined 14:43 cdg joined 14:47 khw joined
viki OK, makese sense. 14:48
yoleaux 10:32Z <lizmat> viki: in general, if .comb returns a List, it's because it already has all elements it could find already reified: not much point turning a .List into a .Seq then. Although it would be simple enough to do if we want that because of interface consistency
14:48 BenGoldberg left 14:49 pierre_ left 14:50 espadrine left 14:55 adu joined
Ulti is there a way to get QAST out of Rakudo as a compilation stage? 14:59
similar to how you can do --stage=MAST
psch --target=ast exists
Ulti K 15:00
psch but that's not a complete dump, but a stringification
Ulti that should be enough to help me out
thanks
psch --target=optimize is the AST that comes out of the optimizer, fwiw
so maybe that's more useful to look at 15:01
15:01 pmurias joined
pmurias Ulti: re tutorial on NQP, you can look a t/nqp in the NQP repo to see what's available 15:02
moritz did somebody mention jnthn++'s NQP and Rakudo Internals course already? 15:07
Ulti yup 15:08
moritz ok :-)
that's the closest we have to a tutorial
15:08 domidumont joined 15:10 adu left
Ulti im starting out doing the lexical @F no one wants but is most like the Perl 5 feature ;P 15:10
15:12 kyclark_ joined 15:13 kyclark_ left
psch hmm, since -F implies -a implies -n, can't you just put the slip result into @_? 15:13
s/slip/split/
15:13 CIAvash left, chris2 left
moritz don't be so pragmatic :-) 15:16
Ulti yup
pmurias @F is ugly so it should stand out better ;)
15:16 cdg left
Ulti lets not lose sight of this only exists for cli -e "programs" though 15:16
matching perl5 and ruby interfaces isn't the worst thing in the world for usability 15:17
psch actually i'm not even sure side-stepping the @_-autoviv thingy is easy enough
Ulti I was thinking about someone specifying a full grammar and you get parsed lines :3 but its just not what anyone would want for dirty cli things
psch huh, that's a neat idea though 15:18
Ulti yeah it means you can have named fields
which would be nice
psch like "perl6 -FMText::CSV -e'...' data.csv"
Ulti also better code reuse
psch which, actually, is something i would like for dirty cli things
Ulti you could specify a grammar class like -M
psch: yup 15:19
I think thats a super nice extension
psch ++Ulti :)
Ulti well I'll see how far I get I'm only one step above cargo culting at the moment 15:20
15:20 cdg joined
Ulti also a very Perl 6 specific magic way of doing CLI one liners 15:20
easily a killer feature 15:21
for command line junkies
15:21 bartolin left
Ulti because suddenly its easy to do file format aware greps 15:21
even if its multi line records
RabidGravy concludes that the half-idea to make COMPOSE work from user code is probably forlorn 15:25
15:25 espadrine joined 15:26 bartolin joined 15:27 xinming left 15:29 CIAvash joined, chris2 joined 15:30 wamba joined
pmurias having a perl6 version of ack would be awesome 15:30
15:31 kalkin-_ left, kalkin- joined, stux|RC-only left
moritz might be a bit slow :( 15:32
15:33 rindolf left 15:34 xinming joined, stux|RC-only joined
TimToady well, we still need to optimize the heck out of regexen and grammaren 15:35
15:35 rindolf joined
RabidGravy what larks 15:36
TimToady now that I'm finally done with the conference season, maybe I can have a look or two at that
viki \o/
pmurias õ 15:42
15:43 domidumont left
dalek c: 96ffea2 | (Zoffix Znet)++ | doc/Type/IO/Handle.pod6:
Document IO::Handle.printf

Remove paragraph about workarounds		 15:45
synopsebot6 Link: doc.perl6.org/type/IO/Handle
c: 27b993f | (Zoffix Znet)++ | doc/Language/5to6-perlfunc.pod6:
Remove printf workaround mention from 5-to-6 doc
 15:49
synopsebot6 Link: doc.perl6.org/language/5to6-perlfunc
15:51 xtt2 joined, AlexDaniel joined 15:55 xinming left 16:03 kyclark_ joined 16:07 xtt2 left 16:08 chinarulezzz left 16:12 Actualeyes left 16:17 CIAvash left 16:23 giraffe left, kyclark_ left 16:27 cdg left, zacts left
dalek : 084abca | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Switch post

Don't wanna write my original idea anymore, gonna write
  "Perl 6 Core Hacking: How, Where, What, and Why" instead.
To drive more people to contribute to core code.
If Web Gods willing, the contributing.perl6.org will be done by the time post publishes.		 16:28
16:29 giraffe joined 16:37 krokib joined 16:38 espadrine left 16:39 domidumont joined
dalek : 97c6c0a | tadzik++ | misc/perl6advent-2016/schedule:
Move myself later
 16:49
tadzik free karma for pre-procrastinating, neat
psch huh, curious that this years schedule doesn't share the same format as previous one 16:50
i am exclusively refering to "include the 25th with 'Christmas!'"
+s 16:51
moritz psch: time to earn your karma, then! 17:02
psch moritz: actually i'm just assuming that christmas has already happened and that's why the 25th isn't listed :)
moritz psch: as long as I can remember, there's been a Christmas each year :-) 17:04
psch ah, the trouble with coopting extant terms for specific events /o\ 17:05
17:09 mscha left 17:15 stux|RC joined, stux|RC left 17:24 BenGoldberg joined 17:27 vendethiel left 17:28 vendethiel joined 17:32 vendethiel left 17:33 lolo78 joined 17:35 zacts joined
lolo78 Hi, does anyone knows why picking up a random number between 1 and 1e5 takes almost 10 seconds in the REPL, while doing the same between 1 and 100_000 is instantaneous: 17:38
> my $start = now; my $c = (1..1e5).pick; say now - $start;
9.97741524
> my $start = now; my $c = (1..100_000).pick; say now - $start;
0
It looks as a bug to me, but I may overlook something.
viki 0.o 17:40
m: my $start = now; my $c = (1..1e5).pick; say now - $start;
camelia rakudo-moar 8ca367: OUTPUT«6.2881222␤»
viki m: my $c = (1..1e5).pick; say now - INIT now; # <--- trick for timing without the need for $start
17:40 lolo78_ joined
camelia rakudo-moar 8ca367: OUTPUT«6.45798636␤» 17:40
viki s: (1..1e5), 'pick', \() 17:41
SourceBaby viki, Sauce is at github.com/rakudo/rakudo/blob/8774...ge.pm#L497
viki m: my $c = (1..1e5).pick: 1; say now - INIT now;
camelia rakudo-moar 8ca367: OUTPUT«13.60358848␤»
viki m: my $c = (1..1e5).eager[1]; say now - INIT now 17:42
camelia rakudo-moar 8ca367: OUTPUT«6.1120095␤»
viki lolo78: would you report it as a performance bug? Just email to [email@hidden.address]
lolo78 viki: OK, thanks, I will. 17:43
viki Thanks.
psch m: (1e0..1e5).pick; say now - INIT now 17:44
camelia rakudo-moar 8ca367: OUTPUT«0.26786054␤»
psch don't mix types on the ends for now :)
17:44 stux|RC joined
psch m: (1..1e5).pick.WHAT.say 17:46
viki Ah..
camelia rakudo-moar 8ca367: OUTPUT«(Int)␤»
psch afaiu, for each iteration we're creating something that fits the LHS of infix:<..>, then converting it to fit the RHS, then converting again to store it in the resulting Seq
viki m: (1..100_000).pick; say now - INIT now
camelia rakudo-moar 8ca367: OUTPUT«0.0016929␤»
viki cool :D
psch fitting the RHS is necessary because we have to check how it cmps
m: (1e0..100_000).pick; say now - INIT now
camelia rakudo-moar 8ca367: OUTPUT«6.1173763␤» 17:47
lolo78 psch: OK, thanks for the tip, it is indeed much better, but still quite low compared to the 1..100_000 range.
psch m: (1..100_000).pick; say now - INIT now
camelia rakudo-moar 8ca367: OUTPUT«0.0017836␤»
psch m: (1e0..100_000e0).pick; say now - INIT now
camelia rakudo-moar 8ca367: OUTPUT«0.272768␤»
AlexDaniel huh
psch lolo78: yes, because you're passing natives that still have to be boxed
m: (1/1..100_000/1).pick; say now - INIT now
camelia rakudo-moar 8ca367: OUTPUT«1.0455477␤»
psch and, well, Rat is slow anyway vOv
AlexDaniel um, actually I don't see any good reason why it would be so slow 17:48
psch lolo78: in any case, i don't see why you'd ever generate a sequence incrementing by the natural unit in floating point natives or Rats
AlexDaniel: Rat members are usually native, aren't they?
err, s/members/Attributes/
we only bail from that if we reach out of int space afair 17:49
AlexDaniel m: say (0..Inf).pick 17:50
camelia rakudo-moar 8ca367: OUTPUT«Cannot .roll from a lazy list␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤»
AlexDaniel m: say (0..NaN).pick
camelia rakudo-moar 8ca367: OUTPUT«(timeout)»
psch m: say 1 < NaN
camelia rakudo-moar 8ca367: OUTPUT«False␤»
psch m: say 1 > NaN
camelia rakudo-moar 8ca367: OUTPUT«False␤»
lolo78 psch: I was really testing with much larger numbers (hundreds of digits) so using for example the 1e200 notation was easier than typing all the digits.
psch lolo78: well, it creates a different datatype
m: say 1e0.WHAT 17:51
camelia rakudo-moar 8ca367: OUTPUT«(Num)␤»
psch m: say 1.WHAT
camelia rakudo-moar 8ca367: OUTPUT«(Int)␤»
psch we clearly need flexible-length uparrow notation in core :o)
AlexDaniel #125336 is kinde relevant
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=125336
AlexDaniel a*
lolo78 psch: yes, I gather that now, I did not think about it when I ran my tests. 17:52
psch lolo78: maybe some kind of 'infix:<!^>(Int $a, Int $b) { $a ** (10 ** $b) }' can alleviate some of your typing woes :) 17:53
...maybe minus the parens because infix:<**> is right-assoc, isn't it 17:54
moritz I'd still keep them for clarity :-)
lolo78 psch: I guess I can store 1e200 in a variable and coeece it to an Int before building the range. 17:55
psch lolo78: that too!
lolo78: mind, the performance problem you mentioned is still one we probably want to look into
AlexDaniel lolo78: so please submit a bug report anyway!
psch because, really, why wouldn't we convert the RHS to the LHS' type instead of converting every iteration value to the RHS' type 17:56
lolo78 psch: thanks a lot for your help and very enlightening comments. Ans I will submit a bug report.
s/and/and/
psch lolo78: please tag it '[PERF]' if you do, and thanks! :) 17:57
lolo78: where 'tag' means 'put that string in the email subject'
lolo78 psch: I will. Thanks.
psch huggable: rakudobug
huggable psch, [email@hidden.address] or use perl6 query on rt.perl.org ; see github.com/rakudo/rakudo/#reporting-bugs
17:59 TEttinger joined 18:01 skids joined 18:04 girafe joined, chris2 left 18:06 lizmat_ joined 18:08 TEttinger left, lizmat left 18:09 chris2 joined
lolo78 viki, psch and AlexDaniel: I've just submitted the performance bug report to [email@hidden.address] Thanks for your help. 18:13
AlexDaniel thanks! 18:14
lolo78 Just as an update: thos works fine: 18:21
> (1..1e5.Int).pick; say now - INIT now;
0.0009998
>
s/thos/this/
18:22 TEttinger joined 18:25 vendethiel joined 18:49 stux|RC-only left, Tonik joined 18:54 labster joined 18:57 chinarulezzz joined
stmuk_ rakudo.org/2016/11/27/announce-raku...e-2016-11/ 19:03
viki \o/ 19:11
lizmat_ stmuk_++ 19:16
19:17 xinming joined 19:20 espadrine joined
stmuk_ I'm always surprised about the number of emails about the MSI 19:23
19:26 lizmat_ is now known as lizmat
RabidGravy strange how "fast queues" improve the reliability of the Audio::Liquidsoap tests 19:27
I think it's the source switch on the default playlist pre-empting the resolution of the request queue 19:29
mst timing problems make all pandas sad 19:32
moritz stmuk_++
m: for 1..2 { .say; NEXT say "NEXT $_" } 19:39
camelia rakudo-moar 8ca367: OUTPUT«1␤NEXT 1␤2␤NEXT 2␤»
moritz m: for 1..2 { .say; NEXT say "NEXT $_"; last }
camelia rakudo-moar 8ca367: OUTPUT«1␤»
19:45 rindolf left 19:47 stevieb left
lizmat moritz: did you expect something else? 19:47
19:56 haircode joined 19:57 lolo78 left 19:58 lolo78_ left 20:04 lolo78 joined 20:08 lolo78_ joined 20:16 domidumont left 20:19 pochi left 20:20 stevieb joined 20:22 adu joined 20:26 pochi joined 20:27 AlexDaniel left 20:28 r3m is now known as rantanplan, rantanplan is now known as r3m 20:40 cschwenz joined 20:50 bjz joined 20:53 lolo78 left, lolo78__ joined, lolo78__ is now known as lolo78 20:54 lolo78_ left 21:10 adu left 21:19 eisen74 joined, FROGGS left 21:26 rindolf joined
dmaestro I'm still curious about the quoting behavior here: 21:36
m: Qw { [ ] { } }.perl
camelia ( no output )
dmaestro m: say Qw { [ ] { } }.perl 21:37
camelia rakudo-moar 8ca367: OUTPUT«("[", "]", "\{}")␤»
dalek c: 6a28ad9 | coke++ | doc/Language/ (2 files):
remove trailing whitespace
 21:38
c: 53719f6 | coke++ | doc/Language/glossary.pod6:
remove unwarranted contraction
synopsebot6 Link: doc.perl6.org/language/glossary
[Coke] we need more volunteers for advent. 21:42
14 empty slots out of 25.
(I am happy to yield day one to someone writing a non-technical annual retrospective. :) 21:43
21:43 bjz left 21:57 rindolf left 22:06 mj41 joined 22:18 stevieb left 22:21 bjz joined
dalek c: 6c9bfc9 | (Daniel Mita)++ | doc/Language/operators.pod6:
Fix links to postcircumfix operators
 22:21
synopsebot6 Link: doc.perl6.org/language/operators
c: 3904632 | RabidGravy++ | doc/Language/operators.pod6:
Merge pull request #1023 from mienaikage/operator-urls

Fix links to postcircumfix operators
synopsebot6 Link: doc.perl6.org/language/operators
22:27 _28_ria joined 22:30 stevieb joined, pmurias left 22:33 bjz left 22:34 cschwenz left 22:40 MasterDuke left 22:44 kurahaupo_ left 22:45 kurahaupo__ joined 22:50 pmurias joined
timotimo Ulti: MAST doesn't come with a stringification, i don't think; the internal mast stuff is not text-like at all, fwiw 22:50
22:51 RabidGravy left
[Coke] anyone mind if I upgrade perl6 on hack? 22:51
(the rakudobrew version)
... hope not, pulling trigger now 22:54
timotimo i'll be fine with it, but i hardly use perl6 on hack anyway 22:55
[Coke] I just need some place that I can run a service on to test out my article.
22:56 imcsk8_ is now known as imcsk8
[Coke] ... but I need to make sure it's at least using the latest version. 22:56
22:56 kurahaupo__ left
timotimo sure thing, hack is good for that 22:56
though of course you're free to build your own rakudo in your own home directort
directory %)
22:56 seatek joined 22:57 kurahaupo_ joined, kurahaupo left
[Coke] it was .06 22:58
timotimo yowsers
23:00 Actualeyes joined 23:04 pierre_ joined 23:05 pierre_ left 23:06 Actualeyes left 23:08 cdg joined 23:10 kurahaupo_ left 23:12 MasterDuke joined
timotimo whoa, don't we usually shorten the rakudo star changelog significantly? 23:14
for the release i mean
timotimo fixes some markup explosiveness 23:15
23:16 BenGoldberg left 23:17 lolo78 left 23:18 ZzZombo_ joined, ZzZombo left
timotimo fixed 23:20
23:24 AlexDaniel joined
seatek smoldering ruins? 23:25
timotimo well ... with some band-aids
seatek :)
timotimo check it out on rakudo.org 23:26
i'm not sure if i want to turn the bulleted list from a + character in each line to actually using html markup
seatek this is the change from .10 to .11 ? 23:30
timotimo no, it's only rakudo star
i'm not sure when the last rakudo star was
seatek ah ok
23:30 Actualeyes joined
timotimo oh 23:30
the last rakudo star release was .10
seatek yeah that wasn't long ago at all 23:31
timotimo yeah
23:31 lukaramu joined
timotimo i'm going to bed, i'm beat :) 23:32
seatek oh my gosh i didn't even think that you could use multiple adverbs
timotimo hm, we should perhaps make extra sure to include a piece about that in the docs
seatek i'll try to find it and mention it if you like 23:33
timotimo thanks!
it's interesting among other things for :delete; like :delete:kv will give you keys and values, and also delete stuff from a hash 23:34
anyway, AFK
23:34 firstdayonthejob left
seatek like a dangerous factory line 23:34
efficient and deadly 23:35
23:40 wamba left 23:58 cdg left 23:59 pmurias left

Powered by Cro and the Raku® Programming Language.

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

Thank you!