[00:24] *** pubwrk left [00:26] *** patrickb left [00:49] https://gist.github.com/FCO/a684907e94c79daa48aa7c748688ce7f [01:07] m: say 41, 42 [01:07] m: 41 [&say] 42 [01:07] rakudo-moar 0b8ede6c4: OUTPUT: «4142␤» [01:07] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤4142␤Useless use of &say in sink context (line 1)␤» [01:07] Any explanation for this one? [01:10] m: quietly 41 [&say] 42 [01:10] rakudo-moar 0b8ede6c4: OUTPUT: «4142␤» [01:19] *** veesh left [01:29] m: $ = 41 [&say] 42 [01:29] rakudo-moar 0b8ede6c4: OUTPUT: «4142␤» [01:29] ZzZombo: ^^ [01:29] I know how to silence it. I don't understand why it happens. [01:29] m: say 41, 42 [01:29] m: 41 [&say] 42 [01:29] rakudo-moar 0b8ede6c4: OUTPUT: «4142␤» [01:29] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤4142␤Useless use of &say in sink context (line 1)␤» [01:29] No difference for me there ^^. [01:30] m: 1 + 1 [01:30] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of "+" in expression "1 + 1" in sink context (line 1)␤» [01:30] being infix? [01:31] does infix implies you should use the return for something? [01:33] *** veesh joined [01:33] *** hythm joined [01:37] *** wildtrees left [01:45] m: sub bla($a) { $a * 2 }; bla 21 [01:45] rakudo-moar 0b8ede6c4: ( no output ) [01:45] m: sub bla($a) is pure { $a * 2 }; bla 21 [01:45] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of "bla 21" in expression "bla 21" in sink context (line 1)␤» [01:46] ZzZombo: one of the reasons for this warning... [01:47] `&say` isn't pure though? [01:48] ZzZombo: here is where your example is dyeing: https://github.com/rakudo/rakudo/blob/master/src/Perl6/Optimizer.nqp#L2705 [01:48] ZzZombo: I know... I was just showing a reason for that error msg... [01:51] ZzZombo: I think it's because of the sigil... [01:51] m: my \sss = &say; 41 [sss] 42 [01:51] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Missing infix inside []␤at :1␤------> 3my \sss = &say; 41 [7⏏5sss] 42␤ expecting any of:␤ bracketed infix␤ infix␤ infix stopper␤» [01:51] :( [01:52] m: &say(4,1) [01:52] rakudo-moar 0b8ede6c4: OUTPUT: «41␤» [01:54] [&say] and + gives the warning in different lines... [01:55] ZzZombo: I've added the line number on each "Useless use of" warn message on Optimizer file... https://www.irccloud.com/pastebin/llTxMzWN/ [01:56] I'll take a look, thanks. [02:56] *** reportable6 left [02:56] *** notable6 left [02:56] *** unicodable6 left [02:56] *** bloatable6 left [02:56] *** greppable6 left [02:56] *** shareable6 left [02:56] *** nativecallable6 left [02:56] *** releasable6 left [02:56] *** benchable6 left [02:56] *** coverable6 left [02:56] *** committable6 left [02:56] *** quotable6 left [02:56] *** squashable6 left [02:56] *** statisfiable6 left [02:56] *** coverable6 joined [02:56] *** unicodable6 joined [02:57] *** reportable6 joined [02:57] *** greppable6 joined [02:57] *** statisfiable6 joined [02:57] *** bloatable6 joined [02:57] *** notable6 joined [02:57] *** nativecallable6 joined [02:58] *** quotable6 joined [02:58] *** squashable6 joined [02:58] *** releasable6 joined [02:58] *** committable6 joined [02:58] *** shareable6 joined [02:59] *** benchable6 joined [03:59] *** benchable6 left [03:59] *** shareable6 left [03:59] *** bloatable6 left [03:59] *** unicodable6 left [03:59] *** notable6 left [03:59] *** committable6 left [03:59] *** notable6 joined [04:00] *** benchable6 joined [04:00] *** wtwt5237 joined [04:00] *** committable6 joined [04:00] *** unicodable6 joined [04:00] *** shareable6 joined [04:02] *** bloatable6 joined [04:02] hello, folks! may I have your suggestions regarding how to best handle math (stat, machine learning, etc) in Raku? I am aware of math::vector and math::matrix and a few other packages. But compared to python, R and matlab, scientific computing in Raku still seems to be preliminary? Would like to know what other people think here. thx [04:06] wtwt5237: Maybe a module for the purpose? [04:09] I am thinking that if there is any possibility, could the raku team build the capability of scientific computing into the core of it? right now, we already have a number of packages. And to do some simple math, we have to load a lot of packages. To do something more dfificult, such as build a ML or DL model, things will be ten times harder. If a [04:09] centralized solution exists (such as R and matlab), it can easily out-compete these languages. my two cents [04:16] wtwt5237: The problem with such approach that same way we could consider built-in AI engine, for example. Or accounting tools. [04:18] R and matlab are specifially targeted at this area, I don't think it makes sense to compete with them. [04:18] And, to my view, installing a module with zef is as simple as it only could be. [04:20] wtwt5237: but if you volunteer and find an elegant way of implemenenting the thing – it could be accepted. [04:21] wtwt5237: more than that, I'm positive it'd be welcomed! [04:21] * vrurg is afk& [04:23] en.. not directly computing with R/matlab would make sense. but i still feel it would help raku catch the current hype of data science, in some way, if we have something that is more convenient [04:27] haha, I am sorry, I sound like someone who is only complaining but not really doing anything. but my day job now is to run a bioinformatics lab in a medical school. i would love to, but really cannot find any time to learn and do any such implementation now. but i do hope raku could be improved to some extent towards that goal, so I can persuade my [04:27] students to learn raku. now they all learn python... [04:30] m: role :: { COMPOSE { $?CLASS.say } } [04:30] rakudo-moar 0b8ede6c4: ( no output ) [04:31] This should leave a NYI message perhaps? [04:39] also [04:39] m: class :: { COMPOSE { $?CLASS.say } } [04:40] rakudo-moar 0b8ede6c4: ( no output ) [04:40] Is this legal at all? [04:40] hm [04:40] m: class C does Int {} [04:40] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Int is not composable, so C cannot compose it␤at :1␤» [04:45] *** wtwt52371 joined [04:48] *** wtwt52371 left [04:49] *** wtwt5237 left [04:54] *** Xliff joined [04:57] \o [04:58] 2019-11-09T11:47:58Z #raku Xliff: if we are going to create aliases for `.^all.grep` and `.^all.map`, why not call they `.^grep` and `.^map`? [04:58] 2019-11-09T11:51:16Z #raku Xliff: and, update is just one of the uses of `.^all.map`... [04:58] Does anyone know if the routines from https://github.com/LLFourn/p6-CompUnit-Util still work? [05:58] *** shareable6 left [05:58] *** benchable6 left [05:58] *** bloatable6 left [05:58] *** notable6 left [05:58] *** committable6 left [05:58] *** unicodable6 left [05:58] *** squashable6 left [05:58] *** coverable6 left [05:58] *** greppable6 left [05:58] *** releasable6 left [05:58] *** statisfiable6 left [05:58] *** nativecallable6 left [05:58] *** reportable6 left [05:58] *** quotable6 left [05:58] *** committable6 joined [05:59] *** greppable6 joined [05:59] *** squashable6 joined [05:59] *** reportable6 joined [05:59] *** releasable6 joined [05:59] *** bloatable6 joined [05:59] *** statisfiable6 joined [06:00] *** unicodable6 joined [06:00] *** shareable6 joined [06:00] *** coverable6 joined [06:00] *** notable6 joined [06:01] *** nativecallable6 joined [06:01] *** benchable6 joined [06:01] *** quotable6 joined [06:22] *** hythm left [06:43] m: class SortedArray does Positional { has @!a handles ;method iterator {@!a.sort.iterator};submethod BUILD(:@!a) {};method new(**@a){return self.bless: :@a}};.say for SortedArray.new: 3,2,1,4; [06:43] m: class SortedArray does Positional { has @!a handles ;has Iterator $!it handles ;submethod BUILD(:$!it, :@!a) {};method new(**@a){return self.bless: :@a, :it(@a.sort.iterator)}};.say for SortedArray.new: 3,2,1,4; [06:43] rakudo-moar 0b8ede6c4: OUTPUT: «1␤2␤3␤4␤» [06:43] rakudo-moar 0b8ede6c4: OUTPUT: «Rakudo::Iterator::ReifiedListIterator.new␤» [06:44] Why does the second case not work? [07:44] *** bloatable6 left [07:44] *** releasable6 left [07:44] *** quotable6 left [07:44] *** benchable6 left [07:44] *** squashable6 left [07:44] *** greppable6 left [07:44] *** coverable6 left [07:44] *** committable6 left [07:44] *** unicodable6 left [07:44] *** nativecallable6 left [07:44] *** statisfiable6 left [07:44] *** reportable6 left [07:44] *** shareable6 left [07:44] *** notable6 left [07:45] *** quotable6 joined [07:45] *** unicodable6 joined [07:45] *** shareable6 joined [07:45] *** squashable6 joined [07:45] *** AlexDaniel left [07:46] *** greppable6 joined [07:46] *** bloatable6 joined [07:46] *** committable6 joined [07:46] *** nativecallable6 joined [07:46] *** coverable6 joined [07:46] *** statisfiable6 joined [07:46] *** releasable6 joined [07:46] *** benchable6 joined [07:46] *** notable6 joined [07:47] *** reportable6 joined [07:54] m: class A {method m {gather {take ^20[++$]}}};A.new.m.say;A.new.m.say for ^2 [07:54] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Unsupported use of $] variable; in Perl 6 please use $*PERL.version or $*PERL.compiler.version␤at :1␤------> 3class A {method m {gather {take ^20[++$]7⏏5}}};A.new.m.say;A.new.m.say for ^2␤» [07:54] m: class A {method m {gather {take ^20[my ++$]}}};A.new.m.say;A.new.m.say for ^2 [07:54] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Malformed my␤at :1␤------> 3class A {method m {gather {take ^20[my7⏏5 ++$]}}};A.new.m.say;A.new.m.say for ^2␤» [07:59] m: class A {method m {take (^20)[$++]}};dd gather {A.new.m};dd gather A.new.m for ^2 [07:59] rakudo-moar 0b8ede6c4: OUTPUT: «(0,).Seq␤(1, 2).Seq␤» [07:59] m: class A {method m {take (^20)[++$]}};dd gather {A.new.m};dd gather A.new.m for ^2 [07:59] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Unsupported use of $] variable; in Perl 6 please use $*PERL.version or $*PERL.compiler.version␤at :1␤------> 3class A {method m {take (^20)[++$]7⏏5}};dd gather {A.new.m};dd gather A.new.m␤» [08:00] Please w0rk. [08:01] So, how can I have separate `gather/take` per instance? [08:02] m: class A {method m {state $a; take (^20)[++$a]}};dd gather {A.new.m};dd gather A.new.m for ^2 [08:02] rakudo-moar 0b8ede6c4: OUTPUT: «(1,).Seq␤(2, 3).Seq␤» [08:03] ^^ ZzZombo [08:07] *** kktt left [08:08] *** kktt joined [08:18] *** xinming_ left [08:19] *** xinming_ joined [08:28] *** ZzZombo_ joined [08:30] *** ZzZombo left [08:30] *** ZzZombo_ is now known as ZzZombo [08:34] m: role :: { $?CLASS.say } [08:34] rakudo-moar 0b8ede6c4: ( no output ) [08:36] m: role R { $?CLASS.say }; class C does R {} [08:36] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Could not instantiate role 'R':␤No such method 'say' for invocant of type 'C'␤at :1␤» [08:37] m: role R { say $?CLASS.^name }; class C does R {} [08:37] rakudo-moar 0b8ede6c4: OUTPUT: «C␤» [08:43] m: .say($_) given &say ∘ &say [08:43] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot resolve caller say(Block+{FakeSignature[Int,Num,Mu]}:D: Block+{FakeSignature[Int,Num,Mu]}:D); none of these signatures match:␤ (Mu: *%_)␤ in block at line 1␤␤» [08:44] m: .say(~$_) given &say ∘ &say [08:44] rakudo-moar 0b8ede6c4: OUTPUT: «Block+{FakeSignature[Int,Num,Mu]} object coerced to string (please use .gist or .perl to do that)␤ in block at line 1␤Cannot resolve caller say(Block+{FakeSignature[Int,Num,Mu]}:D: Str:D); none of these signatures match:␤ (Mu: *%_)␤…» [09:06] *** sena_kun joined [09:15] *** finanalyst joined [09:19] Xliff: I mean, if `$++` is allowed, then `++$` should be too? [09:24] m: say $?FILE.io.lines [09:24] rakudo-moar 0b8ede6c4: OUTPUT: «No such method 'io' for invocant of type 'Str'. Did you mean 'so'?␤ in block at line 1␤␤» [09:24] m: say $?FILE.IO.lines [09:24] rakudo-moar 0b8ede6c4: OUTPUT: «(say $?FILE.IO.lines)␤» [09:28] ZzZombo: It's a parser error, so the use of the variable is a workaround. [09:29] Is it filed? [09:29] No, I don't think so. [09:30] Can you remind me where do bugs go now into? [09:30] https://github.com/rakudo/rakudo/issues [09:35] Thank you! [09:38] *** mid_laptop left [09:57] *** xinming_ left [09:58] *** xinming_ joined [10:14] *** ilogger2 joined [10:25] *** chloekek joined [10:27] * chloekek gonna work on using Raku with Nix again. [10:38] Nix? [10:42] Nix the package manager. [10:42] *** cpan-raku joined [10:42] *** cpan-raku left [10:42] *** cpan-raku joined [10:43] Writing a package for the latest MoarVM, NQP, Rakudo, and adding a function for making packages for Raku distributions. [10:43] Ah! chloekek++ [10:43] Precompiling and PERL6LIB are a pain but I figured out most of it last time. [11:07] *** Altai-man_ joined [11:20] What are the *-j and *-m files in the bin directory of a distribution? [11:40] Ok, that's just different Rakudo backends. [12:01] m: say False but True [12:01] rakudo-moar 0b8ede6c4: OUTPUT: «True␤» [12:02] m: say True but True [12:02] rakudo-moar 0b8ede6c4: OUTPUT: «True␤» [12:02] m: say True but False [12:02] rakudo-moar 0b8ede6c4: OUTPUT: «False␤» [12:02] Why is all this allowed? [12:04] m: 1 does True [12:04] rakudo-moar 0b8ede6c4: ( no output ) [12:04] m: dd 1 does True [12:04] rakudo-moar 0b8ede6c4: OUTPUT: «1␤» [12:04] m: say 1 does True [12:04] rakudo-moar 0b8ede6c4: OUTPUT: «1␤» [12:05] m: say so 1 does False [12:05] rakudo-moar 0b8ede6c4: OUTPUT: «False␤Some exceptions were thrown in END blocks:␤Unhandled exception: No exception handler located for warn␤ at :1 (/home/camelia/rakudo-m-inst-2/share/perl6/runtime/CORE.c.setting.moarvm:)␤ from SETTING::src/core.c/control.pm6:190…» [12:05] Whoops! [12:05] m: say so 1 but False [12:05] rakudo-moar 0b8ede6c4: OUTPUT: «False␤» [12:05] m: say 1 but False [12:05] rakudo-moar 0b8ede6c4: OUTPUT: «1␤» [12:06] So this allows to make a C++ equivalent of `#define true false` in Raku. [12:07] m: say so True but False [12:07] rakudo-moar 0b8ede6c4: OUTPUT: «False␤» [12:07] In the whole program as well, BTW. [12:07] ZzZombo: you can "but" a value of an enum on any object [12:08] I figured as much, but why?? [12:08] it will override the method with the name of the enum for that object and return the given value [12:08] because it can be very handy ? [12:09] like returning 0 from a routine, but it still being accepted as True ? [12:09] m: say 1 but 2 [12:09] rakudo-moar 0b8ede6c4: OUTPUT: «1␤» [12:09] ?? [12:09] 2 is not an enum [12:09] But what does that even do to the `1`? [12:10] m: say (1 but 2).Int [12:10] rakudo-moar 0b8ede6c4: OUTPUT: «2␤» [12:10] it overrides the Int method... interesting :-) [12:10] guess it's more general than I thought :) [12:11] Why did it display as 1 then? [12:11] because "say" calls .gist on the value [12:11] ($a but $b) adds a method $b.^name to $a which returns $b I thought [12:11] m: say so 1 does False [12:11] rakudo-moar 0b8ede6c4: OUTPUT: «False␤Some exceptions were thrown in END blocks:␤Unhandled exception: No exception handler located for warn␤ at :1 (/home/camelia/rakudo-m-inst-2/share/perl6/runtime/CORE.c.setting.moarvm:)␤ from SETTING::src/core.c/control.pm6:190…» [12:11] What's up with this though? [12:12] m: say ("foo" but 2).Int [12:12] rakudo-moar 0b8ede6c4: OUTPUT: «2␤» [12:12] m: say +("foo" but 2) [12:12] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5foo' (indicated by ⏏)␤ in block at line 1␤␤» [12:12] m: say ("foo" but 2).Numeric [12:12] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5foo' (indicated by ⏏)␤ in block at line 1␤␤» [12:12] I think prefix:<+> calls .Numeric [12:15] m: say 'str' but Nil [12:15] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot mix in non-composable type Any into object of type Str␤ in block at line 1␤␤» [12:15] m: say 'str' but 123 [12:15] rakudo-moar 0b8ede6c4: OUTPUT: «str␤» [12:15] m: say 'str' but m// [12:15] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Null regex not allowed␤at :1␤------> 3say 'str' but m/7⏏5/␤» [12:15] m: say 'str' but m/asd/ [12:15] rakudo-moar 0b8ede6c4: OUTPUT: «Use of uninitialized value of type Any in string context.␤Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.␤ in block at line 1␤Cannot mix in non-composable type Any into object of type Str…» [12:16] m: say 'str' but -> {} [12:16] rakudo-moar 0b8ede6c4: OUTPUT: «str␤» [12:16] m: say -> {} but &say [12:16] rakudo-moar 0b8ede6c4: OUTPUT: «-> { #`(Block+{}|59422520) ... }␤» [12:17] m: &say.wrap: &exit [12:17] rakudo-moar 0b8ede6c4: ( no output ) [12:18] m: say 'str' but Failure [12:18] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot mix in non-composable type Failure into object of type Str␤ in block at line 1␤␤» [12:18] m: say 'str' but Any [12:18] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot mix in non-composable type Any into object of type Str␤ in block at line 1␤␤» [12:18] m: say 'str' but Array [12:18] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot mix in non-composable type Array into object of type Str␤ in block at line 1␤␤» [12:18] m: use NativeCall;say 'str' but Pointer [12:18] rakudo-moar 0b8ede6c4: OUTPUT: «Cannot mix in non-composable type NativeCall::Types::Pointer into object of type Str␤ in block at line 1␤␤» [12:19] So what defines whether a type can be composed as a role like that? [12:28] *** thundergnat joined [12:30] ZzZombo: but is useful when you want to mix a role into an _instance_. If you want to apply it to a whole class, use does. [12:34] ZzZombo: See http://rosettacode.org/wiki/Peano_curve#Perl_6 - Mixes the Lindenmeyer role into the string 'L'. [12:35] Sorry that I keep bringing up Rosettacode. I'm an admin there so am very familiar with the code examples on there. [12:41] thundergnat: Ooh! I was trying to think of that name for the last day or so. Thanks for reminding me! :) [12:41] Question tho...for the $peano... what are the 'L' and 'R' for? [12:42] *** pubwrk joined [12:43] Xliff: they don't mean anything specific, they just are keys to point to the "expansion" applied by the Lindenmeyer system. [12:43] Ah. [12:43] They could be anything. [12:44] They just happen to be named L and R in this example. [12:47] m: say .file, ':', .line with &infix: [12:47] rakudo-moar 0b8ede6c4: OUTPUT: «SETTING::src/core.c/operators.pm6:65␤» [12:47] Still trying to figure out how the rules are set up. [12:48] ZzZombo: according to that ^ you can always do it with .DEFINITE things. For indefinites, like type objects, that apparently goes into metaobject land which I don't understand [12:48] m: say 123 but Pointer.new [12:48] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Undeclared name:␤ Pointer used at line 1␤␤» [12:48] m: use NativeCall; say 123 but Pointer.new [12:48] rakudo-moar 0b8ede6c4: OUTPUT: «123␤» [12:49] Xliff: for 'L' and 'R', replace where they occur in the string by their expansion. All other characters get processed when you comb through the string. [12:51] No. I HOW it expands. I just didn't know about that particular way to invoke a role with a parameter. [12:52] timotimo first taught me about Lindenmayer and I was using tr///! [12:55] Oh. Sorry. It overrides the .succ method and use $var++ to do each step. [12:55] s/HOW/know HOW/ [12:56] The thing that was getting me was "Lindenmayer(%!rules)" [12:56] I didn't know that was a shortcut to initialize the only PUBLIC attribute of a role. Heh [12:59] There's a lot of pretty cool examples of Raku (Perl 6) code on Rosettacode. [12:59] Yep! [12:59] > 1100 different task examples. [13:00] I love RosettaCode [13:01] FWIW, we (The active Perl 6 contributors) have elected to not start changing the name over at least until there is a Rakudo release with a "raku" compiler. [13:01] to cut down on confusion. [13:01] Yeah. That makes sense. [13:03] m: say &say xor &warn [13:03] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤&say␤Useless use of &warn in sink context (line 1)␤» [13:03] At this moment, Perl 6 is one task entry away from being tied for second for the most task entries. (hint, hint if anybody wants to do some Perl 6 coding) [13:04] m: say(&say xor &warn) [13:04] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Unable to parse expression in argument list; couldn't find final ')' (corresponding starter was at line 1)␤at :1␤------> 3say(&say7⏏5 xor &warn)␤ expecting any of:␤ infix␤…» [13:04] What's wrong? [13:05] thundergnat: Where is that listing? [13:05] http://rosettacode.org/wiki/Reports:Tasks_not_implemented_in_Perl_6 [13:06] m: dd &warn [13:06] rakudo-moar 0b8ede6c4: OUTPUT: «Sub warn = proto sub warn (|) {*}␤» [13:09] m: m $a = &say xor &warn;dd $a [13:09] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5===␤Unrecognized regex metacharacter = (must be quoted to match literally)␤at :1␤------> 3m $a7⏏5 = &say xor &warn;dd $a␤Couldn't find terminator $ (corresponding $ was at line 1)␤at :1␤------> 3m $a =7⏏…» [13:09] m: my $a = &say xor &warn;dd $a [13:09] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of &warn in sink context (line 1)␤Sub $a = proto sub say (|) {*}␤» [13:10] WTF [13:10] *** Altai-man_ left [13:11] thundergnat: http://rosettacode.org/wiki/Use_another_language_to_call_a_function --> NativeCall? [13:11] weekly: https://twitter.com/raku_cpan_new [13:11] lizmat, Noted! (weekly) [13:12] m: my $a = &say xor 0;dd $a [13:12] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of constant integer 0 in sink context (line 1)␤Sub $a = proto sub say (|) {*}␤» [13:12] m: my $a = 1 xor 0;dd $a [13:12] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of constant integer 0 in sink context (line 1)␤Int $a = 1␤» [13:12] m: my $a = 1 xor $;dd $a [13:12] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of unnamed $ variable in sink context (line 1)␤Int $a = 1␤» [13:13] m: 1 xor 1 [13:13] rakudo-moar 0b8ede6c4: OUTPUT: «WARNINGS for :␤Useless use of constant integer 1 in sink context (line 1)␤» [13:13] thundergnat: Oops! No, it's the other way around. Want to call rakudo code from another language. [13:13] Xliff: yeah, that is a little tougher. [13:23] Xliff: something with cro maybe ? [13:34] ZzZombo: xor has extremely loose precedence, like and and or. `my $a = &say xor 0` turns out to be `(my $a = &say) xor (0)` which gives you sink context warning for the second expression. [13:37] but I'm very unsure about if it should be looser than parentheses in a function call [13:39] m: my $a = (1 xor $);dd $a [13:39] rakudo-moar 0b8ede6c4: OUTPUT: «Int $a = 1␤» [13:39] hm [13:40] ref: rosetta code. what about having a raku repo for the solutions so anyone can contribute? [13:40] afk& [13:41] tbrowder: It's a wiki, anyone CAN contribute. [13:54] *** chloekek left [13:58] lizmat: What did you mean by Cro? [14:03] Is there a good Perl6 module for parsing HTML, yet? [14:03] I've generally been using Mojo::DOM:from [14:04] Xliff: Gumbo has worked very well for me. A lot of the early crashing problems seem to have been cured. [14:09] Hmmm.... installing it now [14:11] And it's broken... :( [14:12] well... it works for me... [14:12] Heh. Looks like a forking issue. I've cloned it. [14:13] So it's fixed in git and not through zef. Weird. [14:13] *** xinming_ joined [14:14] Xliff: Sigh, yeah I run into that a bunch. Things get fixed on github but a updated version doesn't get pushed to CPAN. [14:21] *** stoned75 joined [14:23] *** lucasb joined [14:30] m: say((0 or 10)); say (0 or 10) [14:30] rakudo-moar 0b8ede6c4: OUTPUT: «10␤10␤» [14:30] m: say(0 or 10) [14:30] rakudo-moar 0b8ede6c4: OUTPUT: «5===SORRY!5=== Error while compiling ␤Unable to parse expression in argument list; couldn't find final ')' (corresponding starter was at line 1)␤at :1␤------> 3say(07⏏5 or 10)␤ expecting any of:␤ infix␤ …» [14:31] yeah, it's a little strange that the parentheses from the function call doesn't work. dunno what are the syntatic limitations.... [14:35] *** chloekek joined [14:38] *** patrickb joined [15:05] m: sub f(Str @a) { say @a }; my @a = ; f @a; [15:05] rakudo-moar 0b8ede6c4: OUTPUT: «Type check failed in binding to parameter '@a'; expected Positional[Str] but got Array ($["a", "b"])␤ in sub f at line 1␤ in block at line 1␤␤» [15:05] 2019-11-06T18:00:21Z #raku-dev <[Coke]> uzl[m] command-tab is mac "switch between windows" [15:07] m: sub f(Str @a) { say @a }; my @a = Array[Str].new('a', 'b'); f @a; [15:07] rakudo-moar 0b8ede6c4: OUTPUT: «Type check failed in binding to parameter '@a'; expected Positional[Str] but got Array ($["a", "b"])␤ in sub f at line 1␤ in block at line 1␤␤» [15:09] m: sub f(Str @a) { say @a }; my Str @a = ; f @a; [15:09] rakudo-moar 0b8ede6c4: OUTPUT: «[a b]␤» [15:12] *** MasterDuke joined [15:18] *** pmurias joined [15:20] thundergnat: Looks like Yahoo has modified its search page to prevent automated searches. [15:21] Oh, but the old form still works. Hah. [15:23] Xliff: yeah, some of the remaining tasks are based on apis that have changed since the task was written. [15:26] Ah, but I'm pulling search results down, now. ;) [15:28] m: my $*a = True; -> $? where CALLER::<$*a> {}() [15:28] m: my $*a = True; -> $? where $*a {}() [15:28] rakudo-moar 0b8ede6c4: OUTPUT: «Constraint type check failed in binding to parameter ''; expected anonymous constraint to be met but got Mu (Mu)␤ in block at line 1␤␤» [15:28] rakudo-moar 0b8ede6c4: ( no output ) [15:28] Why does the first case not work? [15:29] Xliff: https://github.com/FCO/Red/issues/406 [15:30] guys, would you guys mind of helping on a Red's API decision? https://github.com/FCO/Red/issues/406#issue-520615433 [15:31] *** mid_home joined [15:32] m: my $*a = True; -> $? where CALLERS::<$*a> {}() [15:32] rakudo-moar 0b8ede6c4: ( no output ) [15:33] ZzZombo: I suppose the where clause check happens somewhere in the guts, not inside the pointy block. [15:33] https://twitter.com/smokemachine/status/1193552092272766976 [15:33] not directly inside, anyway [15:34] m: my $*a = True; -> $? where DYNAMIC::<$*a> {}() [15:34] rakudo-moar 0b8ede6c4: ( no output ) [15:35] m: my $*a = True; -> $? where DYNAMIC('$*a') {}() [15:35] rakudo-moar 0b8ede6c4: ( no output ) [15:35] More, more! [15:35] Xliff: have you seen the survey? [15:35] m: my $*a = True; dd DYNAMIC:: [15:35] rakudo-moar 0b8ede6c4: OUTPUT: «PseudoStash.new(("!UNIT_MARKER" => !UNIT_MARKER,"\$!" => Nil,"\$*a" => Bool::True,"\$/" => Nil,"\$=finish" => Mu,"\$=pod" => [],"\$?PACKAGE" => GLOBAL,"\$_" => Any,"\$¢" => Nil,"::?PACKAGE" => GLOBAL,:EXPORT(EXPORT),:GLOBALish(GLOBAL)))␤» [15:36] SmokeMachine: Yes. Commented [15:42] Xliff: and I've also added a comment on your PR... there is 1 ")" missing there... [15:42] 'k [16:31] vrurg: what dp you think? https://github.com/FCO/Red/issues/406 [16:35] thundergnat: https://gist.github.com/Xliff/5fb6df9f353c812502ece6bf15e59755 [16:38] p6: grammar G { rule TOP { '(' * ')' } }; G.parse('(()()(()))', actions => G); [16:38] rakudo-moar 0b8ede6c4: OUTPUT: «Too many positionals passed; expected 1 argument but got 2␤ in regex TOP at line 1␤ in block at line 1␤␤» [16:46] ref rosetta code: is it "fair" to use native call on a c or c++ version of, say, an avl tree? [16:46] I don't think it's instructive [16:46] the point of the site is to show how stuff is implemented in different languages [16:47] i see other langs doing similar, but, as moritz says, not as instructive. [16:47] and if for example the task is to compute pi, shelling out the the pi (1) command line tool doesn't show how to calculate pi in Raku [16:50] *** pmurias left [16:51] that's why i asked about a repo, say start out with a translation of another lang's version of a known algorithm, then iteratively improve it with community involvement and rakudoisms before putting it on rosetta [16:52] or throw up a gist and let the shots begin? [16:52] Heh. I just did that. [16:56] Xliff, cool. i've tried p6, er, raku xml a bit and had problems with one of the xml mods, but i found a little workaround. let me look... [17:00] *** jmerelo joined [17:03] *** Voldenet joined [17:03] *** Voldenet left [17:03] *** Voldenet joined [17:08] hm, i see where i use module XML successfully. but you're starting with html, not actual xml... [17:09] Nope. Using Gumbo and they convert that to an XML::Document [17:18] tbrowder: "ref rosetta code: is it "fair" to use native call..." Somewhat discouraged, though, create a nativecall module and push it to the ecosystem and you're gold. [17:19] what probs are you seeing with XML? [17:20] Xliff: ^^^ [17:21] thundergnat: maybe two versions: one native call and one pure raku? [17:21] At least with Gumbo created XML::Document, it's not finding certain nodes by class. [17:22] Having problems extracting the text via XML::Element.contents [17:22] XML::Element.lookfor() should be recursive, but it is returning wrong counts for certain tags. [17:25] I'd be delighted to get task solutions, even if it is calling external libraries. A bunch of the image manipulation task do that, and while it isn't pure Perl, it is instructive about _how_ a task may be approached. [17:25] it would be handy to have some raku translators to-from xml <-> json or yaml or some other non-brain-hurting format [17:26] thundergnat: thnx [17:27] *** pubwrk left [17:31] *** stoned75 left [17:57] *** sena_kun joined [17:58] how do I have printf without "print" part? [17:58] sprintf [17:59] hmm, really... need to check [17:59] docs say "Formats and outputs a string" [17:59] *** pubwrk joined [17:59] docs seem wrong [18:07] Well, it _does_ output a formatted string, it just returns it to the caller rather than printing it to $*OUT. [18:14] I have a question about Red, why are operations like save handled with the model metaclass instead of the class? [18:15] so that custom methods on the model objects are free to be named whatever you like perhaps [18:15]