»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by masak on 12 May 2015.
cognominal class A { method a is export { say 'a'}}; # so I suppose it means I can call it A::a as well as A.a 00:02
makes more sense
psch m: class A { has $.b; method a is export { say $.b }}; import A; a A.new(:b(5))
camelia rakudo-moar fd8c98: OUTPUT«5␤»
00:02 adu left
psch m: class A { has $.b; method a is export { say $.b }}; import A; say &a.signature 00:03
camelia rakudo-moar fd8c98: OUTPUT«(A $: *%_)␤»
00:03 RabidGravy left 00:07 colomon left 00:08 colomon joined
cognominal what is the dollar in the signature? 00:08
psch cognominal: a scalar invocant
m: class A { has $.b; method a is export(A $self: ) { say $self!b }}; import A; say &a.signature 00:09
camelia rakudo-moar fd8c98: OUTPUT«5===SORRY!5=== Error while compiling /tmp/btTGXDIdP6␤Two terms in a row␤at /tmp/btTGXDIdP6:1␤------> 3class A { has $.b; method a is export(A7⏏5 $self: ) { say $self!b }}; import A; sa␤ expecting any of:␤ infix␤ infix…»
psch oh
that's not where the signature goes...
m: class A { has $.b; method a (A $self: ) is export { say $self!b }}; import A; say &a.signature
camelia rakudo-moar fd8c98: OUTPUT«5===SORRY!5=== Error while compiling /tmp/yE4twEl1Gl␤No such private method 'b' for invocant of type 'A'␤at /tmp/yE4twEl1Gl:1␤------> 3hod a (A $self: ) is export { say $self!7⏏5b }}; import A; say &a.signature␤»
psch oh, the param is outside..? 00:10
m: class A { has $.b; method a (A $self: ) is export { say $self.b }}; import A; say &a.signature
camelia rakudo-moar fd8c98: OUTPUT«(A $self: *%_)␤»
psch anyway, there
cognominal psch++ 00:12
00:16 tokuhirom joined
psch ah, no, if it's not literal self it always goes through method dispatch... 00:16
m: class A { has $!b; method a (A $self: ) is export { say self!b; say $self =:= self }}; import A; a A.new(:b(1)) 00:17
camelia rakudo-moar fd8c98: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ci7T9lM3z2␤No such private method 'b' for invocant of type 'A'␤at /tmp/ci7T9lM3z2:1␤------> 3thod a (A $self: ) is export { say self!7⏏5b; say $self =:= self }}; import A; a A.␤»
cognominal m: class A { my &a = method a { say 'a' }; method b { a A }}; A.b # wicked
camelia rakudo-moar fd8c98: OUTPUT«a␤»
psch well, okay, that has me a tad confused...
oh duh
$.a is the same as self.a, but self!a isn't the same as $!a 00:18
konobi is there a version of the 'nix' language in nqp land? 00:19
00:20 tokuhirom left
[Coke] not to my knowledge. 00:34
cognominal m: class A { my &a = method { say 'a' }; my &b = method { say 'b' }; our @a = ( &a, &b ) }; A.@A::a 00:43
camelia rakudo-moar 1051fc: OUTPUT«Invocant requires a type object of type Array, but an object instance was passed. Did you forget a 'multi'?␤ in block <unit> at /tmp/Vr4mqfhqbE:1␤␤»
cognominal m: class A { my &a = method { say 'a' }; my &b = method { say 'b' }; our @a = ( &a, &b ) }; say @A::a
camelia rakudo-moar 1051fc: OUTPUT«[<anon> <anon>]␤»
01:00 tokuhirom_h joined 01:04 tokuhirom_h left
cognominal m: 1.infix:['+'](2) 01:06
camelia rakudo-moar 1051fc: OUTPUT«Method 'infix:<+>' not found for invocant of class 'Int'␤ in block <unit> at /tmp/oXpkTheTnw:1␤␤»
cognominal m: say infix:['+'](1, 2) 01:08
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/F4e_z7kUwW␤Undeclared routine:␤ infix:['+'] used at line 1. Did you mean 'infix:<+|>', 'infix:<+>', 'infix:«+>»', 'infix:«+<»', 'infix:<+^>'?␤␤»
01:08 yqt left
AlexDaniel .u ⚠ 01:10
yoleaux U+26A0 WARNING SIGN [So] (⚠)
sprocket is there something akin to a PERL5LIB environment var for rakudo that i can use to specify alternate library paths? 01:11
cognominal sprocket, I think this is in flux right now. 01:12
sprocket cognominal: what’s the current method? i just need a way to test a module i’m working on, before i package it up
cognominal I think you can use -I when invoking rakudo 01:13
sprocket ah ok
cognominal I see PERL6LIB in the sources 01:14
github.com/rakudo/rakudo/blob/nom/...ore/Inc.pm 01:15
sprocket cognominal: I’ll check it out - thank you!
cognominal m: say 1.:<+>(2) 01:16
camelia rakudo-moar 1051fc: OUTPUT«Cannot find method 'CALL-ME'␤ in block <unit> at /tmp/bCGdBoTIi0:1␤␤»
cognominal at least they all parse :) 01:17
raiph .ask itz_stmuk re: pl6anet stats... Any idea what happened on Oct 21 (flat visits/pages etc but a crazy hit spike)? ... Why are so many hits 404s (70% in Oct!)? 01:23
yoleaux raiph: I'll pass your message to itz_stmuk.
zengargoyle is there a particular reason FIRST { … } can't be used in a while loop? 01:26
but LAST { … } can. 01:27
m: my $x = True; while $x { FIRST { say "first" }; LAST { say "last" }; $x = False; } 01:29
camelia rakudo-moar 1051fc: OUTPUT«last␤»
zengargoyle m: my $x = True; while $x { once { say "first" }; LAST { say "last" }; $x = False; }
camelia rakudo-moar 1051fc: OUTPUT«first␤last␤»
01:29 finanalyst left
psch m: say &infix:['+'](1,2) 01:31
camelia rakudo-moar 1051fc: OUTPUT«3␤»
psch hm, not sure if that's autogen or the other bit...
gfldex m: my $i = 10; while $i-- { FIRST { say 'first' } }; 01:32
camelia ( no output )
gfldex m: my $i = 10; while $i-- { FIRST { say 'first' } say 'alive' };
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/_ychOoa7Lc␤Strange text after block (missing semicolon or comma?)␤at /tmp/_ychOoa7Lc:1␤------> 3= 10; while $i-- { FIRST { say 'first' }7⏏5 say 'alive' };␤ expecting any of:␤ infix␤ …»
gfldex m: my $i = 10; while $i-- { FIRST { say 'first' }; say 'alive' };
camelia rakudo-moar 1051fc: OUTPUT«alive␤alive␤alive␤alive␤alive␤alive␤alive␤alive␤alive␤alive␤»
gfldex zengargoyle: please rakudobug
zengargoyle heh, seems like i'm rakudobuging something every day :P 01:33
01:35 TEttinger joined 01:36 cognominal left
mst design.perl6.org/Differences.html 01:36
is there any chance we can replace 'perl6 has a "real" object system' with something saying it's built into the language rather than provided by libraries or whatever
or if that doc's considered 'out of date, meh' can we put some sort of thing at the top saying that 01:37
01:38 MadcapJake joined
zengargoyle rakudobug'd: [perl #126535] AutoReply: [BUG] - FIRST phaser not firing in while 01:39
synbot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=126535
AlexDaniel .u 🛂 01:47
yoleaux U+1F6C2 PASSPORT CONTROL [So] (🛂)
01:48 aborazmeh joined, aborazmeh left, aborazmeh joined
[Coke] m: say "\c[PASSPORT CONTROL]" ~ "\c[COMBINING DIAERESIS]" 01:51
camelia rakudo-moar 1051fc: OUTPUT«🛂̈␤»
01:53 xpen joined 01:55 laouji joined
AlexDaniel .u 💬 01:55
yoleaux U+1F4AC SPEECH BALLOON [So] (💬)
01:55 laouji left
psch m: say "\c[PASSPORT CONTROL]\c[COMBINING SPEECHBALOON]" # i wonder... 01:56
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/5beV6hPyp4␤Unrecognized character name COMBINING SPEECHBALOON␤at /tmp/5beV6hPyp4:1␤------> 3SSPORT CONTROL]\c[COMBINING SPEECHBALOON7⏏5]" # i wonder...␤»
psch aww unicode, why not
:P
TEttinger m: say "\c[PASSPORT CONTROL]\c[COMBINING SPEECHBALLOON]" 01:57
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/E_CVaKrj52␤Unrecognized character name COMBINING SPEECHBALLOON␤at /tmp/E_CVaKrj52:1␤------> 3SPORT CONTROL]\c[COMBINING SPEECHBALLOON7⏏5]"␤»
TEttinger m: say "\c[PASSPORT CONTROL]\c[COMBINING SPEECH BALLOON]"
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/_z1WSSNTDm␤Unrecognized character name COMBINING SPEECH BALLOON␤at /tmp/_z1WSSNTDm:1␤------> 3PORT CONTROL]\c[COMBINING SPEECH BALLOON7⏏5]"␤»
TEttinger aw
is there a way to search the unicode DB for all matches? 01:58
[Coke] .u SPEECH
yoleaux U+2EC8 CJK RADICAL C-SIMPLIFIED SPEECH [So] (⻈)
U+2F94 KANGXI RADICAL SPEECH [So] (⾔)
U+1F4AC SPEECH BALLOON [So] (💬)
[Coke] .u SPEECHB
yoleaux No characters found
AlexDaniel .u BALOON 01:59
yoleaux No characters found
AlexDaniel oops
.u BALLOON
yoleaux U+2723 FOUR BALLOON-SPOKED ASTERISK [So] (✣)
U+2724 HEAVY FOUR BALLOON-SPOKED ASTERISK [So] (✤)
U+2749 BALLOON-SPOKED ASTERISK [So] (❉)
AlexDaniel so it is limited to just 3
TEttinger .u COMBINING SPEECH 02:01
yoleaux No characters found
TEttinger .u COMBINING BALLOON
yoleaux No characters found
TEttinger .u COMBINING BUBBLE
yoleaux No characters found
TEttinger .u SPEECH BALLOON
yoleaux U+1F4AC SPEECH BALLOON [So] (💬)
TEttinger hm
AlexDaniel there is no combining bubbles, stop it! :D 02:02
psch "\x0".."\xffff" .grep({.uniprop eq 'Mn'}) # get all the marks!
AlexDaniel here's the list of all combining stuff: en.wikipedia.org/wiki/Combining_character 02:03
02:03 adu joined
AlexDaniel 🛂⃠ 02:04
psch ooh, COMBINING HALF-WIDTH LEFT ALIGNED BLOCK 02:05
02:06 sprocket left
psch better goes to bed :S 02:06
02:08 goofyFuzzball joined
diakopter haha 02:15
02:22 risou left
Sgeo Does/will Perl6 have a reasonable FFI? 02:22
02:25 yeahnoob joined
AlexDaniel Sgeo: like Inline::Python? 02:27
zengargoyle TEttinger: are you on a linux box?
TEttinger nope
Sgeo I don't know anything about Inline::Python
Sgeo wants to use an opaque C library
adu opaque?
TEttinger why, zengargoyle?
gfldex Sgeo: see NativeCall 02:28
adu and Inline::C
gfldex design.perl6.org/S21.html
adu and C::Parser
zengargoyle ah, don't know about Win/Mac... i usually zcat /usr/shar/i18n/charmaps/UTF-8.gx | fgrep <whatever>
TEttinger ah ok
zengargoyle other OS might have a file somewhere of characters...
adu zengargoyle: Mac is pretty close to most other unixen 02:29
zengargoyle yeah, i almost count Mac, but don't know file locations.
not sure, think that file is part of iconv 02:30
02:30 bpmedley left
adu zengargoyle: usually if you substitute historical monikers with TitleCaseWords, the directory structure is pretty close 02:30
zengargoyle uh, no locales package it seems. 02:31
adu zengargoyle: like /home is /Users, /usr is /Library and lib/ is Frameworks/ and include/ is Headers/ 02:32
zengargoyle lol
adu zengargoyle: it's exactly the same :P 02:33
zengargoyle: but I think most developer I know install homebrew or macports to smooth out the differences 02:34
zengargoyle yeah, scripting and unix-y app wise i mostly consider them equivalent... it's now 'will probably work on Linux/Mac, don't know about Windows' as long as you can find the package in your package manager. 02:36
and is more likely to work on Mac than Solaris. :P
adu most unix-y people I know who use windows install Cygwin
02:37 BenGoldberg left
adu I've tried Solaris, it's /dev tree is like an alien world 02:37
zengargoyle quite a bit of Solaris at $WORK, but i leave those details to others. it's sorta very slowly being phased out ever since the Oracle took over. 02:41
02:41 goofyFuzzball left
gtodd adu: and if you want, you can even make symlinks to TitleCaseWords and olde fashioned unix to feel more part of the NeXT-generation 02:44
02:47 cognominal joined 02:59 cognominal left 03:05 kaare_ joined 03:10 aborazmeh left 03:12 MadcapJake left, atweiden left 03:18 risou_awy joined, risou_awy is now known as risou
raiph m: sink 1 03:20
camelia ( no output )
03:20 cognominal joined 03:22 tokuhirom joined 03:28 finanalyst joined 03:31 cognominal left 03:32 aborazmeh joined, aborazmeh left, aborazmeh joined 03:37 Actualeyes joined 03:38 xinming left
zengargoyle back in college days i kept my laundry in a NeXT box. too cool. 03:40
03:44 xinming joined 03:54 Oatmeal left 04:00 Oatmeal joined 04:04 khw joined 04:06 kaare_ left, aborazmeh left
dalek kudo-star-daily: 5ec59e7 | coke++ | log/ (9 files):
today (automated commit)
 04:29
04:39 Actualeyes left 04:43 adu left 04:44 skids left 04:45 llfourn joined 04:58 Actualeyes joined 05:04 khw left 05:05 sprocket joined
Hotkeys does perl 6 have an equivalent to the $INPUT_RECORD_SEPARATOR or $RS or $/ variables from perl 5? 05:07
moritz Hotkeys: you can specify the separator during open with nl => "separator" 05:08
TimToady I thought there was a problem with autochomp if you do that 05:09
yoleaux 1 Nov 2015 21:39Z <lizmat> TimToady: I'm having a hard time describing your work on autogen the past week. Could you summarize that in a line of 1,2,3 (or more) for the Perl6 Weekly?
TimToady anyway, at the moment I think slurp.split is faster
05:09 atweiden joined
TimToady hopefully will change in the next week or two 05:09
moritz there's also open(...).split iirc 05:13
which might be less memory hungry
TimToady .tell lizmat There's a sense in which none of the metaoperators are really there, but are abstractions; up till now they've autovivified themselves as needed only when used in a normal operator slot, but now they autovivify when mentioned as nouns or function names too, in any of &foofix:<op>, foofix:<op>, or &[op] now also honor the abstraction 05:15
yoleaux TimToady: I'll pass your message to lizmat.
raiph .tell atweiden It's something like Rats are Rat[64]s -- they overflow to Nums when the denominator of a result overflows 64 bits so consider using Rat[128+]s or FatRats instead 05:23
yoleaux raiph: I'll pass your message to atweiden.
05:23 AlexDaniel left 05:31 sprocket left
atweiden raiph: thanks, didn't know about Rat64 05:32
yoleaux 05:23Z <raiph> atweiden: It's something like Rats are Rat[64]s -- they overflow to Nums when the denominator of a result overflows 64 bits so consider using Rat[128+]s or FatRats instead
atweiden any thoughts on Rat64 vs rat64 native types?
moritz rats are composite types; there's not much point having a native one 05:33
TimToady we had specced a native form that has a numerator twice as big as the denominator, since there's not much use in limiting numbers to be close to 1 when when the denom is huge 05:34
so a rat64 would actually have an int128 for its numerator 05:35
05:42 [Sno] left
zengargoyle m: say "ア つ.ぐ".match(/<:Kana :Hira [.\h]>+/); 05:45
camelia rakudo-moar 1051fc: OUTPUT«｢ア つ.ぐ｣␤»
zengargoyle m: say "ア つ.ぐ".match(/<:Kana :Hira [.]>+/);
camelia rakudo-moar 1051fc: OUTPUT«｢ア｣␤»
arnsholt zengargoyle: [...] doesn't introduce a character class in Perl 6, FYI =)
It's a non-capturing group like (?:) in Perl 5 05:46
TimToady it's inside <>
zengargoyle insise < >
TimToady but you need +
m: say "ア つ.ぐ".match(/<:Kana+:Hira+[.]>+/);
camelia rakudo-moar 1051fc: OUTPUT«｢ア｣␤»
TimToady m: say "ア つ.ぐ".match(/<:Kana+:Hira+[.\h]>+/);
camelia rakudo-moar 1051fc: OUTPUT«｢ア つ.ぐ｣␤»
zengargoyle m: say "ア つ.ぐ".match(/<:Kana :Hira [.\h]>+/); 05:47
camelia rakudo-moar 1051fc: OUTPUT«｢ア つ.ぐ｣␤»
arnsholt Oh, derp. It would appear that reading comprehension deteriorates as thesis writing intensifies >.<
zengargoyle m: say "ア つ.ぐ".match(/<:Kana+:Hira+[.]>+ % ' '/);
camelia rakudo-moar 1051fc: OUTPUT«｢ア つ｣␤»
zengargoyle why is [.] not matching the '.' 05:48
TimToady that...shouldn't work work without +
05:48 finanalyst left
zengargoyle tried both ways and decided + didn't make a difference and thought maybe it was the default... 05:49
TimToady m: say "ア つ.ぐ".match(/<[\w] [.\h]>+/); 05:50
camelia rakudo-moar 1051fc: OUTPUT«｢ア つ.ぐ｣␤»
zengargoyle m: say "つ.ぐ".match/<:Hira+[.]>+/ 05:51
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at /tmp/J8YuEQDvI1:1␤------> 3say "つ.ぐ".match/<:Hira+[.]>+/7⏏5<EOL>␤Regex not terminated.␤at /tmp/J8YuEQDvI1:1␤------> 3say "つ.ぐ".match/<:Hira+[.]>+/7⏏5<EOL>␤Unable to parse regex; …»
TimToady eep, + isn't supposed to be default
zengargoyle m: say "つ.ぐ".match/<:Hira+[.]>+/);
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5===␤Null regex not allowed␤at /tmp/xag7STumLQ:1␤------> 3say "つ.ぐ".match/<:Hira+[.]>+/7⏏5);␤Unable to parse regex; couldn't find final '/'␤at /tmp/xag7STumLQ:1␤------> 3say "つ.ぐ".match/<:Hira+[.]>+/7⏏5);␤ …»
zengargoyle m: say "つ.ぐ".match(/<:Hira+[.]>+/);
camelia rakudo-moar 1051fc: OUTPUT«｢つ.ぐ｣␤»
zengargoyle m: say "つ.ぐ つ.ぐ".match(/<:Hira+[.]>+ % ' '/); 05:52
camelia rakudo-moar 1051fc: OUTPUT«｢つ｣␤»
05:53 raiph left
zengargoyle m: "つ.ぐ つ.ぐ" ~~ m :g /<:Hira+[.]>+ % ' '/; say @(); 05:56
camelia rakudo-moar 1051fc: OUTPUT«(｢つ｣ ｢.｣ ｢ぐ つ｣ ｢.｣ ｢ぐ｣)␤»
zengargoyle wishes linenoise handled wide characters. 05:59
atweiden is it Rat128? getting invalid typename. 06:08
maybe i'll try rat64 first, but is it implemented? 06:09
m: my rat64 $r = 0.0 06:16
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5===␤Type 'rat64' is not declared␤at /tmp/4JyC2ay_s2:1␤------> 3my rat647⏏5 $r = 0.0␤Malformed my␤at /tmp/4JyC2ay_s2:1␤------> 3my7⏏5 rat64 $r = 0.0␤␤»
atweiden m: my rat32 $r = 0.0
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5===␤Type 'rat32' is not declared␤at /tmp/nL33VHIKZU:1␤------> 3my rat327⏏5 $r = 0.0␤Malformed my␤at /tmp/nL33VHIKZU:1␤------> 3my7⏏5 rat32 $r = 0.0␤␤»
atweiden m: my Rat64 $r = 0.0 06:17
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5===␤Type 'Rat64' is not declared␤at /tmp/8ajKDsPbz7:1␤------> 3my Rat647⏏5 $r = 0.0␤Malformed my␤at /tmp/8ajKDsPbz7:1␤------> 3my7⏏5 Rat64 $r = 0.0␤␤»
atweiden how do you declare one of these larger Rats?
TimToady no, all we have is Rat and FatRat so far 06:31
06:37 jevin_ joined 06:42 quester joined
quester p6: my $i = 65536; say $i; $i *=2; say $i 06:42
camelia rakudo-moar 1051fc: OUTPUT«65536␤131072␤»
quester p6: my int $i = 65536; say $i; $i *=2; say $i 06:58
camelia rakudo-moar 1051fc: OUTPUT«65536␤131072␤»
06:58 quester left 07:06 diana_olhovik_ joined 07:12 baest_ is now known as baest 07:27 firstdayonthejob joined 07:32 [Sno] joined 07:38 stux|RC-only joined 07:41 abraxxa joined 07:42 RabidGravy joined 07:45 darutoko joined
RabidGravy marnin 07:46
07:46 abraxxa left 07:55 colomon left 07:56 colomon joined 08:00 abraxxa joined 08:09 kjs_ joined 08:17 _mg_ joined 08:20 atweiden left, ely-se joined 08:28 firstdayonthejob left, Humbedooh left 08:30 Humbedooh joined
lizmat good *, #perl6! 08:33
yoleaux 05:15Z <TimToady> lizmat: There's a sense in which none of the metaoperators are really there, but are abstractions; up till now they've autovivified themselves as needed only when used in a normal operator slot, but now they autovivify when mentioned as nouns or function names too, in any of &foofix:<op>, foofix:<op>, or &[op] now also honor the abstraction
lizmat .tell TimToady Thanks for the quote!
yoleaux lizmat: I'll pass your message to TimToady.
nine lizmat: do you actually understand this? 08:34
dalek kudo/nom: ee87d79 | lizmat++ | t/01-sanity/53-transpose.t:
Fixup TRANSPOSE tests after changes
08:35 zakharyas joined
lizmat I think so :-) 08:35
nine lizmat: then I look forward to your explanation in the weekly :)
moritz does 08:36
llfourn is there any fun that can be had with CompUnit interface that is worth documenting yet?
lizmat well, my interpretation is that if you say [+] @array, it is actually auto-vivified because they are needed
moritz there are infinitely many operators that you can create from basic operators and meta operators
lizmat whereas if you say my $a = &[+], they weren't yet
nine llfourn: there's much fun in completely refactoring CompUnit and friends ;) 08:37
llfourn takes that as a no
moritz for example from infix:<+>, you can create X[ZR+] or so
so we don't create them up-front, but on demand
llfourn nine: good luck with that :)
moritz that used to work for calls, but not for references to those operators
now it works for references too
m: say &infix:<Z+>.signature.perl 08:38
camelia rakudo-moar 1051fc: OUTPUT«:(+ is raw)␤»
moritz star-m: say &infix:<Z+>.signature.perl
camelia star-m 2015.09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Tpk56BPwpu␤Undeclared routine:␤ &infix:<Z+> used at line 1. Did you mean '&infix:<∖>', '&infix:<lt>', '&infix:<eq>', '&infix:<>=>', '&infix:<=>>'?␤␤»
nine moritz: thanks! Crystal clear now 08:39
lizmat nine: what moritz said
moritz: can I use your exegesis in the P6W ?
TimToady notes that infix:<+>(1,2) is technically a call too, but didn't work before 08:40
yoleaux 08:33Z <lizmat> TimToady: Thanks for the quote!
TimToady also, &[+] is an existing operator
lizmat TimToady: ah. duh :-) 08:41
moritz lizmat: sure 08:42
star-m: say infix:<+>(1,2)
camelia star-m 2015.09: OUTPUT«3␤»
moritz star-m: say infix:<Z+>(1,2)
camelia star-m 2015.09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/6rTbHU4LNg␤Undeclared routine:␤ infix:<Z+> used at line 1. Did you mean 'infix:<∖>', 'infix:<lt>', 'infix:<eq>', 'infix:<>=>', 'infix:<=>>'?␤␤»
mrf * #perl6 08:44
RabidGravy rtp
er
TimToady ah, yeah, infix:<Z+> didn't work before, duh 08:46
TimToady should go to bed soon...
08:46 pdcawley joined
TimToady btw, I'm okay with /x ** ^10/ but not okay with /x ** $x/, 'cuz I wanna keep the {}-less form static, and because people will get confused by whether a subscript is allowed there 08:47
as for /x ** ^*/, that's pretty darn useless, and confusing to look at, given that both ^ and * are normal regex metachars 08:48
mrf TimToady: That that seems fair. 08:49
TimToady mrf++ for the ^10 form though
mrf I only added ^* as it seems to keep consistent with the use of ^ as a range
TimToady (though calling '^' min seems a bit weird)
TimToady does wonder if people will think ^10 means 0..10 though, when it should mean 0..9 08:50
ShimmerFairy I would expect 0..9, as it is everywhere else :) 08:51
RabidGravy depends on what you think "up to 10" means ;=) 08:52
mrf hmm I actually expected ^10 to be the same as 0..10. I had clearly missed that it was 0..9
llfourn I exptected 0..10 tbh but 0..10 makes more sense when you think about it. If for ^10 iterated 11 times that would be confusing.
lizmat most people should think of it as 10 times
RabidGravy yes
lizmat m: .say "foo" for ^10
camelia rakudo-moar 1051fc: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CccAJuZGHc␤Two terms in a row␤at /tmp/CccAJuZGHc:1␤------> 3.say7⏏5 "foo" for ^10␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier…»
lizmat m: say "foo" for ^10 08:53
camelia rakudo-moar 1051fc: OUTPUT«foo␤foo␤foo␤foo␤foo␤foo␤foo␤foo␤foo␤foo␤»
ShimmerFairy m: say ^10
camelia rakudo-moar 1051fc: OUTPUT«0..^10␤»
ShimmerFairy m: say (^10).list 08:54
camelia rakudo-moar 1051fc: OUTPUT«(0 1 2 3 4 5 6 7 8 9)␤»
ShimmerFairy ^ that's what I think of when I think of the upto syntax. So I'd expect ** ^10 to mean ** 0..9
mrf I might update the Range type docs to make it clearer that ^ is exclusive of the endpoint
lizmat m: say + ^10
camelia rakudo-moar 1051fc: OUTPUT«10␤»
mrf Also I suspect that might mean there is a bug in my NQP patch 08:55
m: say "foo" ~~ /o ** 0..2/ 08:57
camelia rakudo-moar ee87d7: OUTPUT«｢｣␤»
ShimmerFairy The discovery of /a ** ^10/ not working was because ^10 is another way to write a Range which was accidentally not covered. Deviating from that (by suddenly pretending it's 0..10 instead of 0..^10) would be a bug in my eyes.
mrf m: say "foo" ~~ /o ** 0..^2/
camelia rakudo-moar ee87d7: OUTPUT«5===SORRY!5=== Error while compiling /tmp/sI2dB5PMRF␤Malformed Range␤at /tmp/sI2dB5PMRF:1␤------> 3say "foo" ~~ /o ** 0..7⏏5^2/␤»
mrf ShimmerFairy: Agreed. I believe my patch would treat it as 0..10 not 0..^10 which would be erroneous. 08:58
ShimmerFairy m: say "AAAA" ~~ /A ** {0..^4}/ # this is also a bug in how regexes handle Ranges, though IIRC I came across it a while back :) 08:59
camelia rakudo-moar ee87d7: OUTPUT«｢AAAA｣␤»
Hotkeys <lizmat> m: say + ^10
did you mean
m: say [+] ^10
camelia rakudo-moar ee87d7: OUTPUT«45␤»
lizmat no, just showing the number of elems in ^10 09:00
Hotkeys oh
TimToady m: say "AAAAAAAA" ~~ /A ** {0..^4}/
camelia rakudo-moar ee87d7: OUTPUT«｢AAAA｣␤»
TimToady okay, it's just an off-by-one there due to not considering the exclusion
but definitely a bug
dalek kudo/nom: 63f6fc9 | lizmat++ | src/core/Str.pm:
Also use TRANSPOSE for tr/x// case
 09:01
ShimmerFairy TimToady: if it's using .bounds on the returned Range there, could that possibly be a more general issue with .bounds not considering exclusion?
m: say (1..^4).bounds
camelia rakudo-moar ee87d7: OUTPUT«(1 4)␤»
09:01 espadrine joined
TimToady we need an intbounds methods or so 09:01
mrf I will take a look at it and see if I am able to fix it to allow exclusionary ranges
RabidGravy I actually that that the { ... } bits in regexes got handed off to perl proper to determine the value 09:02
timotimo mrf: sorry for not spotting that off-by-one :S
lizmat TimToady: I've been wanting that for a while, will add that now
ShimmerFairy RabidGravy: yeah, so my first guess is that Range.bounds is being passed to the regex engine without correction :)
mrf timotimo: no worries. You have been super helpful and given me mre than enough support 09:03
timotimo++
TimToady lizmat: note the code for .elems already has the logic
mrf psch++ # also helping
timotimo yay
lizmat TimToady: I'm going one step further and making $!excludes-min/max native ints internally 09:04
mrf timotimo: Considering I thought I would be incabable of doing anything. You got me to the point where I could submit a patch. That is an awesome thing.
Is ^10 a synonym for "10 iterations" then?
Timbus 'upto 10'
creates a list from 0..9 09:05
TimToady ^10 is just short for 0 ..^ 10
mrf Timbus: Sadly upto ten is subjective. Some people would be inclusive and some wouldn't
Timbus luckily, it looks a lot like the long form
mrf TimToady: That is probably a better answer.
TimToady but understanding that, "upto" is how it's pronounced :) 09:06
mrf Timbus: It does but I caught me out. I was expecting it to be equivelant of 0..10.
TimToady: ok
I suppose that is consistent with its use in a full range.. 09:07
TimToady it's our answer to python's [:10] tomfoolery 09:08
mrf TimToady: I will remove the ^* option if your would prefer. Though I am now curious as to what "upto infinity" actually means :D 09:10
ShimmerFairy mrf: probably the * quantifier :P 09:11
Timbus lucky it stops before you get there eh
mrf ShimmerFairy: likely yes
TimToady it's the opposite of "to infinity and beyond"
mrf :D
ShimmerFairy
.oO( / a ** ^* / --> "ERROR: You used three asterisks, along with a caret, where just one is enough. Please unjam your shift and 8 keys as needed and try again" :P )
 09:12
mrf ShimmerFairy: Thats not actually a terrible unhelpful error. 09:13
timotimo ShimmerFairy: that error message should introspect the currently active keyboard layout
shift-8 on my system is „, for example
ShimmerFairy timotimo: of course, but just think of the ungodly mess of dependencies in that :P 09:14
timotimo: it also doesn't account for the * key on my numpad, fwiw :)
timotimo very problematic!
ShimmerFairy mrf: aside from the casual tone of my fake message, I wonder if ** ^* and/or ** ^Inf would be worth noticing. 09:17
RabidGravy m: module Foo { sub foo() is export { } }; import Foo; say &foo::BLURB # what is BLURB refering to here? 09:26
camelia rakudo-moar 63f6fc: OUTPUT«(Any)␤»
mrf ShimmerFairy: If we want to say that ^* and ^Inf are likely a sign that the user is doing something unnecessary then it seems fair to warn them about it. 09:28
RabidGravy "Perl 6: a million characters of helpful error message with some useful code attached" 09:30
09:36 azawawi joined
azawawi hi 09:36
moritz RabidGravy: uhm, nothing? 09:37
RabidGravy eh?
moritz RabidGravy: do you think it should?
m: say &sin::NOTACOSINE
camelia rakudo-moar 63f6fc: OUTPUT«(Any)␤»
RabidGravy no I stumbled on it by accident and just wondered if it was something, like something defined in the code or something 09:38
timotimo azawawi: when you wrote "i'm on it" on the .bat launcher issue, it sounded like "i'm going to fix this", now you're asking "any updates on this one?", i'm confused 09:41
nine timotimo: I don't get it. I changed the line in the optimizer to "if $optype eq 'p6for' && $op.ann('context') eq 'sink' {" yet it still changes both parts of the QAST::Want or none. 09:43
timotimo ah, hehe. 09:44
nine timotimo: aah... I guess that's because of $past.annotate('statement_level', -> { $sinkee.annotate('context', 'sink') });
timotimo as an optimization, we often share QAST objects
azawawi timotimo: i fixed it and nobody merged it... So i was asking when it will be committed... sorry if i was not clear. I will update it :)
nine So if the loop is in sink context, both parts are annotated with sink? 09:45