»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo:, alpha:, pugs:, std:, or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by lichtkind on 5 March 2010.
thowe it seems uncommon though. 00:00
lue I personally put [ if thisandthat { \n code here... \n } ]
pausenclown i find myself being very differential in this matter
i would like a 1 line non blocked version of if 00:02
snarkyboojum thowe: I tend to find I align my style with whatever I'm contributing to, e.g. the Perl 6 code in Rakudo, or what other guys are working on like masak++ :) That seems to feed back into my everyday coding which is kinda interesting
lue (basically, I consider the { part of the line (whether it's saying if, while, etc.), and I let } line up with the block.)
pausenclown: that makes clarity suffer, so that will likely never happen.
pausenclown its fine in Javascript. 00:03
snarkyboojum well, you have statement if (condition) :) 00:04
lue yes, there's that.
rakudo: my $true = True; say "hello" if $true
pausenclown true. but where do i put the fucking else? =)
p6eval rakudo b9aaa3: OUTPUT«hello␤»
snarkyboojum oh you didn't mention else :)
in which case you have a ternary op :P
lue for if/else, use [conditon]??[success]!![failure] 00:05
pausenclown i think the ternary is only for when it has a lvalue
diakopter std: foo: sub bar() { }; goto bar; 00:06
p6eval std 30379: OUTPUT«ok 00:01 108m␤»
pausenclown not just for side effects
lue rakudo: my $true = True; $true ?? say "hi" !! say "boo"; say "hi" if $true else say "boo"; 00:09
p6eval rakudo b9aaa3: OUTPUT«Confused at line 11, near "say \"hi\" i"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
lue rakudo: my $true = True; $true ?? say "hi" !! say "boo";
p6eval rakudo b9aaa3: OUTPUT«hi␤»
lue rakudo: my $true = True; say "hi" if $true else say "boo"; 00:10
p6eval rakudo b9aaa3: OUTPUT«Confused at line 11, near "say \"hi\" i"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
colomon rakudo: rakudo: my $true = True; say $true ?? "hi" !! "boo"; 00:11
p6eval rakudo b9aaa3: OUTPUT«Confused at line 11, near "rakudo: my"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
colomon rakudo: my $true = True; say $true ?? "hi" !! "boo";
p6eval rakudo b9aaa3: OUTPUT«hi␤»
pausenclown rakudo: say Bool::True 00:12
p6eval rakudo b9aaa3: OUTPUT«1␤»
pausenclown rakudo: say Bool::True.WHAT
p6eval rakudo b9aaa3: OUTPUT«Bool()␤»
pausenclown rakudo: say Bool::True + 2 00:13
p6eval rakudo b9aaa3: OUTPUT«3␤»
pausenclown rakudo: say 2 + -(Bool::True)
p6eval rakudo b9aaa3: OUTPUT«1␤»
pausenclown rakudo: say 1 == Bool::True 00:14
p6eval rakudo b9aaa3: OUTPUT«1␤»
pausenclown rakudo: my Bool $x; $x=1 00:15
p6eval rakudo b9aaa3: OUTPUT«Type check failed for assignment␤current instr.: '&die' pc 17293 (src/builtins/Junction.pir:404)␤»
pausenclown rakudo: my Bool $x; $x=Bool::True
p6eval rakudo b9aaa3: ( no output )
pausenclown mmh
not being able to assign 1|0 to Bool seems unperlish 00:16
lue you didn't say anything :)
snarkyboojum rakudo: say Bool::True ~~ Bool
p6eval rakudo b9aaa3: OUTPUT«1␤»
snarkyboojum rakudo: say Bool::True ~~ Int
p6eval rakudo b9aaa3: OUTPUT«0␤»
snarkyboojum unperl6ish perhaps :) 00:17
pausenclown no. i mean like 00:18
rakudo: my Bool $x; $x=1
p6eval rakudo b9aaa3: OUTPUT«Type check failed for assignment␤current instr.: '&die' pc 17293 (src/builtins/Junction.pir:404)␤»
snarkyboojum well
rakudo: my $x = True; $x = 1;
p6eval rakudo b9aaa3: ( no output )
snarkyboojum rakudo: my $x = True; $x.WHAT.say; $x = 1; $x.WHAT.say; 00:19
p6eval rakudo b9aaa3: OUTPUT«Bool()␤Int()␤»
snarkyboojum makes sense to me
pausenclown rakudo: my Bool $x; $x=:bool(1)
p6eval rakudo b9aaa3: OUTPUT«Type check failed for assignment␤current instr.: '&die' pc 17293 (src/builtins/Junction.pir:404)␤»
snarkyboojum more importantly 00:20
rakudo: say True ~~ 1
p6eval rakudo b9aaa3: OUTPUT«1␤»
snarkyboojum rakudo: say False ~~ 1
p6eval rakudo b9aaa3: OUTPUT«0␤»
lue I wonder...
rakudo: my Bool $x = True; $x ~ 1;
pausenclown <-- pillow listening 00:21
p6eval rakudo b9aaa3: ( no output )
lue rakudo: my Bool $x = False; $x ~ 1; say $x
p6eval rakudo b9aaa3: OUTPUT«0␤»
lue so close :)
colomon why are you making the string "01" and then throwing it away? 00:23
lue I was assuming ~ was smart assign, like ~~ is smart match :) 00:25
snarkyboojum hehe - or smart concatenate? :P
colomon ~ is string concat 00:26
lue lemme guess, there is no smart assign :/ 00:27
colomon There is no smart assign.
lue then at least let me assign 1 and 0 to a variable of type Bool :) 00:28
snarkyboojum but 1 and 0 aren't of type Bool :P
lue rakudo: say "hello" if true; say "snarkyboojum" if 1; 00:29
p6eval rakudo b9aaa3: OUTPUT«Could not find sub &true␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
lue rakudo: say "hello" if True; say "snarkyboojum" if 1;
p6eval rakudo b9aaa3: OUTPUT«hello␤snarkyboojum␤»
colomon say 1.Bool.WHAT
rakudo: say 1.Bool.WHAT
lue QED
p6eval rakudo b9aaa3: OUTPUT«Bool()␤»
colomon rakudo: my Bool $x = False; $x = 1.Bool; say $x 00:30
p6eval rakudo b9aaa3: OUTPUT«1␤»
snarkyboojum what colomon said :)
rakudo: say "snarkyboojum" if 45 00:31
p6eval rakudo b9aaa3: OUTPUT«snarkyboojum␤»
snarkyboojum 45 isn't of type Bool either
lue how it typically works is 0 is false, and !0 is true.
colomon oh, hey, there's an operator for .Bool
rakudo: my Bool $x = False; $x = ?1; say $x 00:32
p6eval rakudo b9aaa3: OUTPUT«1␤»
colomon rakudo: my Bool $x = False; $x = ?0; say $x
p6eval rakudo b9aaa3: OUTPUT«0␤»
lue :O
snarkyboojum also
rakudo: say False ~~ 0
p6eval rakudo b9aaa3: OUTPUT«1␤»
snarkyboojum rakudo: say 0 ~~ False
p6eval rakudo b9aaa3: OUTPUT«0␤»
diakopter perlesque: my int $b; loop (my int $a=1;$a<5;$a+=1) { $b = $a; loop (my int $c = $b; $c < 9; $c+=$a) { System::Console.WriteLine($c) } }; System::Console.WriteLine($b) 00:33
p6eval perlesque: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤2␤4␤6␤8␤3␤6␤4␤8␤4␤»
lue maybe that's a precendence bug? 00:34
diakopter rakudo: my Int $b; loop (my Int $a=1;$a<5;$a+=1) { $b = $a; loop (my Int $c = $b; $c < 9; $c+=$a) { say($c) } }; say($b)
p6eval rakudo b9aaa3: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤2␤4␤6␤8␤3␤6␤4␤8␤4␤»
snarkyboojum or it's smart matching in Bool context in the first one but Int in the other (if that's the correct terminology) 00:35
lue rakudo: say (False ~~ 0).WHAT 00:36
p6eval rakudo b9aaa3: OUTPUT«Bool()␤»
lue rakudo: say (0 ~~ False).WHAT
p6eval rakudo b9aaa3: OUTPUT«Bool()␤»
snarkyboojum rakudo: say 0 ~~ +False 00:37
p6eval rakudo b9aaa3: OUTPUT«1␤»
snarkyboojum rakudo: say ?0 ~~ False
p6eval rakudo b9aaa3: OUTPUT«0␤»
snarkyboojum rakudo: say (?0).WHAT; say False.WHAT 00:38
p6eval rakudo b9aaa3: OUTPUT«Bool()␤Bool()␤»
snarkyboojum rakudo: say (?0) ~~ False 00:39
p6eval rakudo b9aaa3: OUTPUT«0␤»
snarkyboojum weird
diakopter ok, time for labels
deferring parameters/arguments for another day
:D
TimToady: what does this mean:
std: foo: sub bar() { }; goto bar;
p6eval std 30379: OUTPUT«ok 00:01 110m␤»
lue hopefully, you're talking about labels for perlesque, and not the whole of P6 ? 00:40
diakopter P6
P6 has labels/goto
perlesque should too, if it wants to be an efficient assembly language 00:41
snarkyboojum colomon: is it a bug that (?0) doesn't ~~ False, or am I misunderstanding something?
diakopter so the compiler writer isn't limited to the control structures built into the language
snarkyboojum heh oh dear 00:42
diakopter ?
colomon rakudo: say False ~~ False
p6eval rakudo b9aaa3: OUTPUT«0␤»
colomon ~~ False always returns false, and ~~ True always returns true.
lue is running. He hates binary spaghetti /o\ 00:43
colomon At least, that's how Rakudo appears to implement things.
snarkyboojum ok - I don't understand something :)
snarkyboojum false can't match anything can it :) 00:43
d'oh 00:44
rakudo: say (?0) ~~ Bool
p6eval rakudo b9aaa3: OUTPUT«1␤»
snarkyboojum ;)
snarkyboojum is a bit slow since cutting out coffee of a morning
colomon I dunno, I'm not sure why it's done that way.] 00:45
snarkyboojum this reminds me of 00:46
rakudo: say False ~~ True 00:47
p6eval rakudo b9aaa3: OUTPUT«1␤»
colomon rakudo: say Mu ~~ True
p6eval rakudo b9aaa3: OUTPUT«Nominal type check failed for parameter '$topic'; expected Any but got Mu instead␤current instr.: 'perl6;Bool;ACCEPTS' pc 362491 (src/gen/core.pir:26828)␤»
snarkyboojum True always matches, False never does
colomon hmmm.... so is Mu failure sensible or a bug? 00:48
lue rakudo: say 3 is 3
p6eval rakudo b9aaa3: OUTPUT«Confused at line 11, near "say 3 is 3"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
lue (i was hoping 'is' existed :/) 00:49
snarkyboojum rakudo: given False { when True { say "matched" } }
p6eval rakudo b9aaa3: OUTPUT«matched␤»
lue rakudo: say False == False 00:51
p6eval rakudo b9aaa3: OUTPUT«1␤»
lue rakudo: say False ~~ False
p6eval rakudo b9aaa3: OUTPUT«0␤»
lue so, "smart" match is telling me false isn't false?
colomon Any True ~~ True (parsewarn) 00:52
Any False ~~ False match (parsewarn)
looks like it is supposed to generate a warning?
snarkyboojum lue: perlcabal.org/syn/S03.html#line_3705
I think it's working like this
"In any case, if you try to smartmatch with ~~ or when, it will recognize True or False syntactically and warn you that it won't do what you expect." 00:53
colomon++ 00:54
TimToady diakopter: it means call the goto function with the result of the bar function as its first argument.
diakopter o_O 00:55
oh.
heh
TimToady goto is just a function syntactically
which is why a forward goto requires quotes
lue it should tell me that smart match is so smart it knows false isn't false :/
diakopter std: foo: sub bar() { }; goto foo; 00:56
p6eval std 30379: OUTPUT«ok 00:01 108m␤»
diakopter hm, perlesque is going to support the subset of those goto function calls that don't have parens, and that have bare label names as the first argument ;) 00:57
colomon std: say False ~~ False
p6eval std 30379: OUTPUT«ok 00:01 109m␤»
diakopter and no forward gotos
snarkyboojum rakudo: sub A { ... }; A().WHAT.say
p6eval rakudo b9aaa3: OUTPUT«Failure()␤»
snarkyboojum rakudo: sub A { }; A().WHAT.say
diakopter actually no
p6eval rakudo b9aaa3: OUTPUT«Null PMC access in find_method('WHAT')␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
diakopter forward gotos are fine
snarkyboojum Null PMC access! :) 00:58
diakopter without quotes tho
TimToady not in Perl 6 they aren't fine 00:58
well, hmm, maybe they could be
std: goto foo; foo: 00:59
p6eval std 30379: OUTPUT«===SORRY!===␤Bogus statement at /tmp/ePo8L4SvBq line 1 (EOF):␤------> goto foo; foo:⏏<EOL>␤Illegally post-declared type:␤ 'foo' used at line 1␤Parse failed␤FAILED 00:01 107m␤»
TimToady std: goto foo; foo:;
p6eval std 30379: OUTPUT«===SORRY!===␤Illegal use of colon as invocant marker at /tmp/GOPGIGm6LT line 1:␤------> goto foo; foo:⏏;␤Undeclared routine:␤ 'foo' used at line 1,1␤Check failed␤FAILED 00:01 108m␤»
TimToady std: goto foo; foo: ;
p6eval std 30379: OUTPUT«===SORRY!===␤Illegally post-declared type:␤ 'foo' used at line 1␤Check failed␤FAILED 00:01 106m␤»
TimToady probably not a good idea 01:00
since the 'foo' is officially parsing as a listop
though in the case of goto, it generally doesn't matter
could probably be forced to work 01:01
lue False isn't false, gotos exist... A bad day for me and P6 :) 01:04
TimToady it isn't that false isn't false, it's that smart matching is differently smart than you. :) 01:05
diakopter quietfanatic: o_O 01:07
lue I'm fine with false not false, but I'm not sure many others will be :)
In fact, I'd probably place False ~~ False as some sort of smart anti-theft code. 01:08
sorear lue: you seem to be under the impression (~~) is supposed to be an equivalence relation 01:13
that's not at all what it does
smart match is a pattern matching DSL 01:14
like regex for any data type
True is U, False is ø
quietfanatic diakopter: O _ o 01:19
lue DSL? (you're saying it's a bit like Regex For Dummies (An operator for the rest of us!™)) 01:22
diakopter perlesque: my int $a = 2; foo: $a += 1; if ($a < 20) { goto foo }; System::Console.WriteLine($a) 01:24
p6eval perlesque: OUTPUT«20␤»
diakopter perlesque: my int $a = 2; foo: $a += 1; if ($a < 10000000) { goto foo }; System::Console.WriteLine($a)
p6eval perlesque: OUTPUT«10000000␤»
diakopter perlesque: my int $a = 2; foo: $a += 1; if ($a < 1000000000) { goto foo }; System::Console.WriteLine($a) 01:25
p6eval perlesque: OUTPUT«1000000000␤»
quietfanatic System::Console::WriteLine?
diakopter CLR .net/mono builtins
quietfanatic Maybe you should shorten that to SCW
or, like say.
diakopter :P
diakopter that's for Perl 6 01:26
this is perlesque
perlesque is burlesque. and assemblish
and werdy
quietfanatic And System::Console::WriteLn-ish
That's hard to say. 01:27
I'd say you need to write some builtin functions.
diakopter but.. I'm trying to discourage humans from typing it
quietfanatic That's saydistic.
diakopter it's supposed to be a ... target language, you see.
TimToady wait, why are you typing it?
diakopter climbed back on the mushroom and nibbled a bit. 01:28
perlesque: goto blah; 01:29
p6eval perlesque: OUTPUT«␤Unhandled Exception: System.InvalidOperationException: Label blah not marked␤at TriAxis.RunSharp.CodeGen.Complete () <0x000db>␤at TriAxis.RunSharp.RoutineGen`1.TriAxis.RunSharp.IDelayedCompletion.Complete () <0x0003b>␤at TriAxis.RunSharp.TypeGen.FlushCompletionQueue ()
..<0x0003c>␤a…
diakopter perlesque: 3; goto blah; sub foo() { blah: 4 }; 01:30
p6eval perlesque: OUTPUT«␤Unhandled Exception: System.InvalidOperationException: Label blah not marked␤at TriAxis.RunSharp.CodeGen.Complete () <0x000db>␤at TriAxis.RunSharp.RoutineGen`1.TriAxis.RunSharp.IDelayedCompletion.Complete () <0x0003b>␤at TriAxis.RunSharp.TypeGen.FlushCompletionQueue ()
..<0x0003c>␤a…
diakopter perlesque: 3; sub foo() { blah: 4 }; goto blah
p6eval perlesque: OUTPUT«␤Unhandled Exception: System.InvalidOperationException: Label blah not marked␤at TriAxis.RunSharp.CodeGen.Complete () <0x000db>␤at TriAxis.RunSharp.RoutineGen`1.TriAxis.RunSharp.IDelayedCompletion.Complete () <0x0003b>␤at TriAxis.RunSharp.TypeGen.FlushCompletionQueue ()
..<0x0003c>␤a…
diakopter 3-SATisfied 01:31
lue
.oO(diakopter: I need a 1up, sorry)
diakopter hmm. it would be a late commute if I went in now. 01:32
lue afk 01:51
spinclad from the backscroll, /me likes [isso]<-=[isit]=->[isnot] for a one-line conditional <-= =-> 01:59
spinclad diakopter: i want to suggest making calls be gotos with arguments, but i haven't got a clean syntax to suggest yet... 02:05
spinclad perlesque: sub foo($ret, $a) { my $b = $a + 1; $ret($b) }; foo(1f, 1); 1f($v): System::Console.WriteLn($v); 02:08
p6eval perlesque: OUTPUT«parsefail␤»
diakopter no parameters yet
nor invocations of scalars 02:09
but a sub can be stored in a scalar now, I think 02:10
perlesque: my sub $foo = sub blah() { 444 }; System::Console.WriteLine($foo);
p6eval perlesque: OUTPUT«Sprixel.Runtime.sub␤» 02:11
diakopter it's actually implicitly typed
don't ask how.
it's some bottom-up way of doing type inference
sorear Sprixel? 02:12
spinclad the only tricky thing i do above is make the 1f label take arguments, and it seems fairly clean
diakopter sprixel, an anagram of perlsix
the engine/runtime that hosts perlesque
spinclad (if you like continuation-passing style) 02:13
diakopter heh.
don't you dare dare me.
sorear I thought sprixel was an implementation in its own right
diakopter it will be, eventually
sorear also, I thought perlesque was an engine
diakopter perlesque is what I'm calling theh assembly language 02:14
think of it as PIR, but with fewer line breaks
and with native types. 02:15
sorear so basically, NQP
diakopter actually I think next perlesque needs compact arrays
nqp doesn't have native types or compact arrays tho
spinclad yeah, arrays could help with args/params 02:16
diakopter spinclad: the reason I said "don't tempt me to add primitives/syntax for continuations" is that sprixel supports them.
just as powerfully as scheme.
spinclad well, there you go then. i just think they would make sense broken out syntactically in a language that's meant to be completely low-level and only for compilers to talk to. 02:18
sorear diakopter: nqp is getting native types next month 02:19
and nqp does have compact arrays
spinclad but... anyway, i haven't quite gotten to suggesting it, so you may utterly disregard my blather here 02:22
sorear nqp: my @arr := pir::new__ps('FixedIntegerArray'); pir::assign__vpi(@arr, 50*1024*1024); say $arr[50000000] 02:25
p6eval nqp: OUTPUT«Confused at line 1, near "say $arr[5"␤current instr.: 'parrot;HLL;Grammar;panic' pc 500 (src/cheats/hll-grammar.pir:197)␤»
sorear nqp: my @arr := pir::new__ps('FixedIntegerArray'); pir::assign__vpi(@arr, 50*1024*1024); say($arr[50000000])
p6eval nqp: OUTPUT«Symbol '$arr' not predeclared in <anonymous>␤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 137 (compilers/pct/src/PCT/HLLCompiler.pir:101)␤»
sorear nqp: my @arr := pir::new__ps('FixedIntegerArray'); pir::assign__vpi(@arr, 50*1024*1024); say(@arr[50000000])
p6eval nqp: OUTPUT«0␤»
sorear nqp: my @arr := pir::new__ps('FixedIntegerArray'); pir::assign__vpi(@arr, 50*1024*1024); @arr[50000000] = 42; say(@arr[50000000]) 02:26
p6eval nqp: OUTPUT«Assignment ("=") not supported in NQP, use ":=" instead at line 1, near " 42; say(@"␤current instr.: 'parrot;HLL;Grammar;panic' pc 500 (src/cheats/hll-grammar.pir:197)␤»
sorear nqp: my @arr := pir::new__ps('FixedIntegerArray'); pir::assign__vpi(@arr, 50*1024*1024); @arr[50000000] := 42; say(@arr[50000000])
p6eval nqp: OUTPUT«42␤»
sorear nqp: my @arr; @arr[50000000] := 42; say(@arr[50000000])
p6eval nqp: OUTPUT«42␤»
diakopter sorear: oh, by compact arrays, I meant arrays that are sequential in memory, and store native types. What do you mean by compact arrays? 02:38
sorear arrays which store native types directly, without indirection through a polymorphic object 02:39
which is what the built-in FixedIntegerArray and related classes do
the last example was a non-compact array; I was trying to crash p6eval-nqp by allocating 50 million polymorphic objects
(but it didn't work because the object array stores nulls directly, not via a Null object) 02:40
diakopter ok, but how could nqp have compact arrays without native types?
sorear the runtime which underlies NQP has native types 02:41
NQP does not yet have the syntactic support for them, but it can use objects composed of them fine
diakopter oh
sorear adding native types to NQP is mostly a question of yak shaving the AST engine to add boxing and unboxing nodes wherever necessary 02:42
diakopter how do you know nqp is getting native types nxt month
sorear because it's on my todo list 02:43
I might fail
diakopter oh; :)
yeah, boxing/unboxing
the compiler that underlies perlesque (RunSharp) does that for me
(adds the boxing/unboxing CIL instructions where needed) 02:44
diakopter perlesque things remaining: sub scaler invocations (named sub invocations work), parameters/arguments/return, next/last, while/until/unless/dowhile, class declarations, struct declarations, arrays (and other generic types such as hash[T]) 02:48
exposing the compiler API to perlesque code in a better way
diakopter string eval 02:49
diakopter contextuals 02:49
diakopter I can't think of anything else at the moment 02:51
parameters/arguments/return is the only one that's not straightforward to me in my mind (that is, I haven't fully thought through how to do it)
I was certain it would work at one point :) 02:52
but I've forgotten.
(how)
diakopter stops rambling.
diakopter perlesque: my int $a = 4; if ($a > 5) { goto bar }; System::Console.WriteLine(32342342); bar: 5; 02:56
p6eval perlesque: OUTPUT«32342342␤»
diakopter perlesque: my int $a = 6; if ($a > 5) { goto bar }; System::Console.WriteLine(32342342); bar: 5;
p6eval perlesque: ( no output )
diakopter :)
thowe Can an array hold more than one type of object? 03:43
sorear sure
thowe OK, because the wikibook seems to avoid showing an array with a mix of types 03:44
if "type" is the right word
sorear well, a type is really just a set of values 03:44
and an array must have all of its elements drawn from the same set
but there's nothing to stop that set from being 'Any' 03:45
or 'Mu' which also includes junctions
thowe so you can't have @a = $b, @c, %d
sorear my @a; is shorthand for my @a of Mu
$b, @c, and %d are all instances of subclasses of Any
they can all fit into an Array of Any
thowe so, do you need to do anything special to say it is an array of Any? 03:46
sorear no
sorear Any is a subclass of Mu 03:46
thowe but you can say that the array can only contain integers if you want? 03:47
sorear yes
thowe what does that syntax look like?
sorear my int @a iirc
thowe Interesting. Thanks.
sorear rakudo: my Int @a 03:48
p6eval rakudo b9aaa3: ( no output )
sorear rakudo doesn't implement native types yet; you have to use the wrapper class Int
like java.lang.Integer
pugssvn r30380 | lwall++ | [STD] warn on attempts to smartmatch with True or False 05:45
moritz_ good morning 06:19
snarkyboojum guten moren moritz_ 06:20
morgen :|
TimToady std: False ~~ True 06:37
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against True always matches; if you mean to test the topic for␤ truthiness, please use :so or *.so or ?* instead at /tmp/mUr1traw2g line 1:␤------> False ~~ True⏏<EOL>␤ok 00:01 109m␤»
TimToady std: True ~~ False
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against False always fails; if you mean to test the topic for␤ truthiness, please use :!so or *.not or !* instead at /tmp/Y8v3HvGcYp line 1:␤------> True ~~ False⏏<EOL>␤ok 00:01 107m␤»
snarkyboojum TimToady - that's awesome :) 06:46
std: False ~~ False 06:47
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against False always fails; if you mean to test the topic for␤ truthiness, please use :!so or *.not or !* instead at /tmp/ofoAmDhSQt line 1:␤------> False ~~ False⏏<EOL>␤ok 00:01 107m␤»
snarkyboojum I get a broken link to svn.pugscode.org/pugs/src/perl6/STD.pm, I take it, it's been renamed to STD.pm6? 06:50
moritz_ yes
snarkyboojum: where is that link?
snarkyboojum good-o
sorear we should put the official highlighted version of STD.pm up
snarkyboojum mortiz_: on my website :)
sorear and link to that
moritz_ sorear: it would be best to have a page with some general information about STD.pm
snarkyboojum the link to STD on the perl 6 wiki is b0rken - www.perlfoundation.org/perl6/index.cgi?std_pm 06:57
and perl6.org/ ? 06:59
yep - and perl6.org
snarkyboojum a quick google shows it's linked in a number of places :) 07:02
snarkyboojum method dumbsmart ($litbool) heh - TimToady++ 07:06
sorear std: given (False) { when (False) {} } 07:07
p6eval std 30380: OUTPUT«ok 00:01 109m␤»
TimToady parens defeat the check currently 07:08
sorear std: given (False) { when False {} } 07:09
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against False always fails; if you mean to test the topic for␤ truthiness, please use :!so or *.not or !* instead at /tmp/TZqkkRqugG line 1:␤------> given (False) { when False⏏ {} }␤ok 00:01 107m␤»
sorear sorry, p5think
snarkyboojum std: given True { when (True) {} } 07:11
p6eval std 30380: OUTPUT«ok 00:01 107m␤»
snarkyboojum std: given True { when True {} }
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against True always matches; if you mean to test the topic for␤ truthiness, please use :so or *.so or ?* instead at /tmp/wjX2RKYmn1 line 1:␤------> given True { when True⏏ {} }␤ok 00:01 107m␤»
snarkyboojum cool
moritz_ std: given Bool::True { } 07:15
p6eval std 30380: OUTPUT«ok 00:01 107m␤»
TimToady it never tests the given
moritz_ oh right 07:16
std: when Bool::True { }
p6eval std 30380: OUTPUT«ok 00:01 107m␤»
snarkyboojum std: when True {}
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against True always matches; if you mean to test the topic for␤ truthiness, please use :so or *.so or ?* instead at /tmp/CmhVcji8A2 line 1:␤------> when True⏏ {}␤ok 00:01 107m␤»
TimToady there are many ways to defeat it, until we get constant folding 07:17
snarkyboojum just looks at 'True' and 'False' too
fun
still awesomerized :)
TimToady possible to get it to do false positive too 07:18
sorear std: constant True := 5; False ~~ True; 07:19
p6eval std 30380: OUTPUT«===SORRY!===␤Malformed constant at /tmp/bsFvwceWx1 line 1:␤------> constant True :⏏= 5; False ~~ True;␤ expecting any of:␤ coloncircumfix␤ signature␤Parse failed␤FAILED 00:01 107m␤»
sorear std: constant True = 5; False ~~ True;
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against True always matches; if you mean to test the topic for␤ truthiness, please use :so or *.so or ?* instead at /tmp/kgvGigyy4X line 1:␤------> constant True = 5; False ~~ True⏏;␤ok 00:01 107m␤»
sorear like that? 07:20
TimToady std: 0 ~~ False.WHAT 07:21
p6eval std 30380: OUTPUT«Potential difficulties:␤ Smartmatch against False always fails; if you mean to test the topic for␤ truthiness, please use :!so or *.not or !* instead at /tmp/IG1au86u8C line 1:␤------> 0 ~~ False⏏.WHAT␤ok 00:01 107m␤»
snarkyboojum I'm sure lue will still be less perplexed :) 07:26
moritz_: so who can update perl6.org to fix the link to STD.pm? 07:29
moritz_ snarkyboojum: everybody with access to the pugs svn 07:30
snarkyboojum ah, so it's already been updated 07:33
moritz_: how about the wiki? 07:34
pugssvn r30381 | moritz++ | [util] the S32 .html files were not generated in the S32/ folder, so try to fix this 07:36
moritz_ snarkyboojum: everybody - it's a wiki :-)
snarkyboojum :O
snarkyboojum moritz_: I should work out how to get a login then :P 07:46
IllvilJa o
pugssvn r30382 | moritz++ | [scripts] fix non-globbing in command line arguments 07:47
Teratogen we need to get Perl 6 out the door, folks
get people in the industry using it
for major projects 07:48
we need a production version of Perl 6 1.0
mberends Teratogen: stop repeating yourself, or we will ... 07:48
moritz_ or I did. 07:49
mathw what was this 'we'? 07:58
snarkyboojum what a funny guy
mathw And good localtime to you 07:59
mathw has tea, and a pleasant morning-after-aikido feeling
moritz_ colomon: I think/hope I fixed the remaining issues with HTML updates of the synopsis, and the ones I randomly checked are up to date 08:15
mathw \o/ 08:17
masak oh hai, #perl6 08:19
mberends oh hai, masak 08:20
frettled Hello, Dr. Morning, I presume.
snarkyboojum masak o/ 08:21
masak snarkyboojum: got a moment for Yapsi sometime during the morning-evening? :) 08:22
snarkyboojum masak: later this evening my time - yes for sure 08:23
masak \o/
snarkyboojum :)
masak wishes there were a established word 'prenoon' in English
mathw 'morning' usually suffices for most people 08:24
mberends masak: could you look at a proto bug that has me stumped? 08:25
masak sure. 08:27
do I just need to run it, or is there any particula incantation that does the stumping? 08:29
mberends :-) it fails to update project state after installing a project, claiming that %!project-state is readonly. Ecosystem.pm6 line 73
for example, './proto install svg' triggers the error 08:30
masak trajs it
mberends most projects do actually get installed, which is kinda useful 08:31
masak \o/
I'm so glad you're digging into this.
moritz_ mberends: that syntax worries me...
%!project-state{$project}<state> = $state;
masak why? 08:32
moritz_ mberends: hashes don't autoviv
in rakudo, at least
oh wait
that's handled above...
moritz_ shuts up
masak my proto dies earlier than that. 08:35
masak "Method 'exists' not found for invocant of class 'Proxy'" 08:36
masak glares angrily at the Proxy class in Rakudo
Proxy: you're of no particular use whatsoever, are you?
masak :) 08:37
masak error occurs on line 95 in Ecosystem.pm6 08:38
so %!project-info{$project} doesn't exist for me.
moritz_ same here 08:39
mberends that may be because the file does not exist at the expected location
spinclad masak: 'forenoon' has a bit of 19th c. sound to it to my ear, tho i'm sure it's in active use in many parts of Britain proper
masak I do have a 'svg:' in my projects.list, though.
spinclad: thanks! you've made my forenoon! 08:40
mberends the file should be ~/.perl6/proto/projects.state
masak mberends: isn't it supposed to find projects.list in the current directory?
mathw spinclad: I'm pretty sure I've seen 'forenoon' used in old writing. I don't think it's in any modern dialects though. 08:41
masak ah.
mberends masak: yes, projects.list has not yet relocated to ~/.perl6/proto 08:42
masak mberends: then I don't get this: up in the 'new' method, there's no path. it calls the load-project-list, which opens that file. there's no path there either.
just the filename itself all the way.
wouldn't it be looking in the current directory, then?
mberends it would 08:43
mberends blushes
spinclad mathw: i expect it's in occasional use in rural or elder or other conservative bywaters
mberends there are *two* projects.state files here, in ~/proto and ~/.perl6/proto
spinclad but i haven't seen it in recent print either 08:44
masak I don't care where the projects.state is located, as long as projects.list sits in the git-charged proto dir itself, so that it gets updated with proto updates.
by the way, I found it slightly confusing that the methods are called load-project-list when the file is called projects.list 08:45
they diff on an 's'.
mberends good point. consistency++
masak (and we can't change the name of the file) :)
mberends ok, will rename the methods. 08:46
masak while I'm being picky, I haven't decided whether it's a smell that we're using load-projects-list to load projects.state... 08:47
mberends I also plan to add a new method to pull down only the latest projects.list from github using LWP::UserAgent
masak maybe 'load-projects-list' isn't the intended name of that method... 08:48
sounds good.
moritz_ general reminder: tonight 18:30 UTC is #rakudosketch meeting
mberends yes. the projects.state got tacked on afterwards
thanks moritz_
masak *nod*
masak will, once again, be unable to attend #rs 08:49
moritz_ :(
masak yes :( 08:49
next Tuesday should be fine.
spinclad ah.. if i go to bed _now_ i can be up by then. 08:51
masak mberends: I'll be happy to look at your bug when the Proxy.exists oddity is gone.
mberends ok 08:52
spinclad sleep(:time(sleeptime)) 08:54
&
masak at the bottom of friendfeed.com/the-life-scientists/...nt-to-mine people are worried about the availability of libs when Perl 6 "comes out". 09:37
this tells us two things: a CPAN bridge would be very, very good. and good native libs wouldn't be so bad either. 09:38
the more we can get working for Rakudo Star, the better.
let's go! www.success.co.il/knowledge/images/...acroix.jpg :) 09:39
moritz_ encourages people not to forget the booko
s/o$//
masak indeed. 09:40
I have two concrete plans. let's say I put them into action this week.
(1) make a chapter about the Poker example
(2) make a chapter with a simple web app 09:41
if I do those two, I'll move over to massive review-editing. 09:43
moritz_ some meta operators are already explained in the chapter on operators 09:44
it's sure worth expanding
masak does (1) now
masak here's the Poker hand example, for those who haven't seen it yet: gist.github.com/244255 09:47
I'm sure it can be updated for Rakudo master a bit.
I'll start with that.
oh noes, named enums! :(
moritz_ :(
masak ok, I'll do this: I'll emulate named enums as much as I can by the latest spec. 09:49
then we can put in the real named enums when the time comes. 09:50
that's a way to 'scout ahead'.
dalek kudo: cb45216 | (Solomon Foster)++ | src/core/Real.pm:
Fix Real.abs and Real.sign to use named invocants.
 09:59
masak colomon++
moritz_ colomon: speaking of which, how is your grant progressing? 10:00
masak rakudo: class A { method foo($bar:) { say $bar.WHAT } }; A.new.foo( :bar(42) )
colomon I need to do a blog post on it. ;)
p6eval rakudo b9aaa3: OUTPUT«Too many positional parameters passed; got 1 but expected 1␤current instr.: 'perl6;A;foo' pc 400 (EVAL_1:171)␤»
masak submits LTA rakudobug
colomon moritz_: very slowly, but surely
moritz_ colomon: I don't expect fast progress during your move anyway :-) 10:01
colomon Things should start picking up now that the moving is over except for unpacking.
moritz_ having moved myself recently I know the extra work, and how slow it returns to 0 afterwards 10:01
masak std: class A { method foo($bar:) { say $bar.WHAT } }; A.new.foo( :bar(42) ) 10:02
colomon I need assistance from jnthn sorting out some bugs with roles.
p6eval std 30382: OUTPUT«ok 00:01 111m␤» 10:03
moritz_ multis from within roles?
colomon moritz_: and more. 10:04
how is cool coming, btw?
masak ready for review and merge, last I heard. 10:06
colomon \o/ 10:08
masak rakudo: package E { our sub a() { 0 }; our sub b() { 1 } }; sub a() { E::a }; sub b() { E::b }; .say for E::a, E::b, a, b # working around lack of named enums
p6eval rakudo b9aaa3: OUTPUT«0␤1␤0␤1␤»
masak lunch & 10:12
pmurias TimToady: viv generates some classes at runtime while it has some pregenerated 10:36
pausenclown g'localtime 10:47
pmurias ruoso: i'm adding caching of the output of STD to mildew (bread actually) 10:54
pausenclown given a somewhat functional XML Parser, what would u consider more important. XPath support, Parsed external entities, or easy access a la XML::Simple?
moritz_ easy access, although I don't like XML::Simple 10:55
maybe XPath would be better
moritz_
.oO( never ask a Perl 6 hacker for advice, 'cause he'll tell you both aye and nay )
 10:57
pmurias aye|nay to be precise 10:58
pausenclown maybe i should ask Grant McLean if he wants to tackle the "easy" part =)
anybody know his irc nick? 10:59
quester_ rakudo: <1 3 4 2>.sort({-$^a}).say 11:04
p6eval rakudo cb4521: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤current instr.: '_block48' pc -1 ((unknown file):-1)␤»
quester_ just wanted to check that was still broken before submitting a patch for it 11:04
pausenclown do we have tools for the new POD yet? 11:31
moritz_ no 11:33
masak pausenclown: I'd very much like XML tooling for Perl 6 that was as fast as libxml. speed is my number one requirement when it comes to XML. 11:39
pausenclown masak: my parser relies on grammars so i'd say its as fast as rakudos IO and the grammar engine. 11:44
masak I admire the effort. I'll probably read through that code somehow, looking for cool ideas. 11:45
pausenclown it's on github
masak URL?
pausenclown github.com/pausenclown/XML--Parser 11:46
moritz_ is still looking for volunteers for the April release of Rakudo 11:47
it's *not* the Rakudo * release
masak :D :/
pausenclown: mind telling me why you choose that... interesting commit comment habit? :) 11:48
pausenclown masak; the docs are really to be trusted.
best to look at the test files 11:49
pausenclown lol. no i dont mind. 11:49
i dont really care in early project stages =)
pausenclown oh. s/really/not really/ 11:50
masak commit comments are probably the first thing I look at when I see a new project. 11:51
this is the largest bla-to-commit-comment ratio I've ever encountered. :) 11:52
moritz_ reduced it a wee bit
masak moritz_++
pausenclown i promise to better myself =)
masak ++pausenclown :)
masak looks at the tests 11:53
masak looks really nice. 11:54
what's that second positional argument to .parse ? 11:55
pausenclown it defines the action class 11:57
masak why a string?
pausenclown like dom, handler
so that the caller doesnt need to create an instance of XML::Parser::Actions::* by himself 11:58
takadonet morning all
masak takadonet: \o
pausenclown but i should probably let him do that if he wants 11:59
masak it doesn't feel very extensible as it is now. 12:00
the new_action method contains four hardcoded cases, with much similarity between them.
pausenclown this is because Foo::Bar::$name doesnt work yet 12:01
masak also, that method does two quite different things.
it sets an attribute and returns a value.
pausenclown not a value. the value it set. 12:03
so it behaves like a classic setter; cant see whats wrong with that. 12:04
masak nevertheless, you could do 'return do given $action { when 'dom' { XML::Parser::Actions::Dom.new( parser => self } #`(other cases); default { die "Unknown XML::Parser::Action" } }' and set the attribute in the parse method. 12:05
saves a lot of repetition, and makes things more orthogonal.
masak in fact, the way you currently do it will not warn if an illegal action string is passed with a .parse call which is not the first .parse call. it will just ignore it and use the last one. 12:07
pugssvn r30383 | pmurias++ | [mildew]
r30383 | enabled caching the parse tree in bread
r30383 | added a -P{STD,STD-cached} option to bread
pausenclown that is correct.
masak++ # glorious beetle hunter 12:08
masak :)
this is when thinking in patters pays off.
s/patters/patterns/
Separation of Concerns is good for the soul, because the points of failure aren't entangled with each other. 12:09
the thing that wasn't separated here was that in new_action, you used the attribute as a temporary vessel for error checking, and that doesn't work when it already had a correct value. 12:10
masak stops pontificating :)
masak anyway, great work. I see there's a lot of effort put into all this. 12:15
pausenclown: hm, github.com/pausenclown/XML--Parser/...Parser.pod says there's a POD error at line 175. 12:17
lisppaste3 pausenclown pasted "better parse" at paste.lisp.org/display/97719 12:19
masak this one breaks DRY... :) 12:21
pausenclown DRY? 12:22
masak though it's arguably better than the last one.
Don't Repeat Yourself.
you list the admissible string values twice.
better to put in a default clause in the given.
pausenclown doesnt default always run? 12:23
masak also, the $action_arg parameter is never used.
moritz_ pausenclown: only when no when-block run
masak pausenclown: all 'when' clauses escape the given if they match.
pausenclown oh.
my bad =)
pausenclown i'm still concerned about how to implement external entity parsing. 12:25
as it involves reading from files from "within" the grammar.
masak rakudo: say "foo" ~~ / f <{ say "bar" ~~ /bar/ }> oo / 12:28
p6eval rakudo cb4521: OUTPUT«bar␤foo␤»
pausenclown .u L B BAR SYMBOL 12:32
phenny U+2114 L B BAR SYMBOL (℔)
pausenclown ty phenny
;-) 12:33
moritz_ .u foo
phenny U+1558 CANADIAN SYLLABICS FOO (ᕘ)
moritz_ .u foobar
phenny moritz_: Sorry, no results for 'foobar'.
moritz_ .u foo bar
phenny moritz_: Sorry, no results for 'foo bar'.
moritz_ well, one can try :-)
pausenclown .u bar 12:34
phenny U+00A6 BROKEN BAR (¦)
pausenclown .u foo .u bar
phenny pausenclown: Sorry, no results
pausenclown Programmer goes into a Bar->(); Bartender says, did u know this arrow is redundant? 12:35
moritz_ an SQL query goes into a bar, looks at two tables and says 'mind if I join you?' 12:37
masak that one made Twitter yesterday :) 12:38
pausenclown lol. 12:39
a friend of mine was arrested at the weekend.
he is somewhat looking like the typical terrorist (beard, muslim, et al). 12:40
in the pub (very crowded) he decicde to make room by climbing the table, yelling Imsch Allah (however that is spelled) and throwing his backback around =)))) 12:41
*backpack 12:42
moritz_ that's soooo old, and on bash.org for years... 12:43
pausenclown damn
worth a try.
masak doesn't get it 12:44
quester_ submits RT #74334: [PATCH] Fix Parcel.sort (fixes the very first example in cloud.github.com/downloads/perl6/bo...10-04.pdf) and Hash.sort
masak quester_++
quester_ Thank you, masak! 12:45
masak no, thank _you_, quester_! :) 12:46
pausenclown switches computers
masak rakudo: my @a = 1,2,3; my ($b, @c) = @a; say @c.perl 12:48
alpha: my @a = 1,2,3; my ($b, @c) = @a; say @c.perl
p6eval rakudo cb4521: OUTPUT«[[2, 3]]␤»
alpha 30e0ed: OUTPUT«[2, 3]␤»
masak submits rakudobug
quester_ It's the least I could do after pmichaud++ implemented sort in ng in February and got it _this_ _close_ to 100% working... 12:49
masak rakudo: my @a = 1,2,3; my (@c) = @a; say @c.perl 12:50
p6eval rakudo cb4521: OUTPUT«[[1, 2, 3]]␤»
masak alpha: my @a = 1,2,3; my (@c) = @a; say @c.perl
p6eval alpha 30e0ed: OUTPUT«[1, 2, 3]␤» 12:51
masak rakudo: my @a = 1,2,3; my @c = @a; say @c.perl
p6eval rakudo cb4521: OUTPUT«[1, 2, 3]␤»
masak wonders if this is related to the issue with double arrays in attributes 12:52
quester_ Good night, all. o/ 13:02
masak quester_: 'night! 13:03
pausenclown_ wtf! pugscode is down- 13:12
no its just the repository 13:13
arnsholt pausenclown_: It's allahu akbar (god is great), btw. Not insh allah (god willing) =) 13:13
(Modulo spelling)
masak there's no correct spelling if you're behind a transcription wall anyway... 13:15
std: subset Quad of Any; sub classify($_) { when Quad {}; when Quad {} }
p6eval std 30383: OUTPUT«ok 00:01 110m␤»
masak rakudo: subset Quad of Any; sub classify($_) { when Quad {}; when Quad {} }
p6eval rakudo cb4521: OUTPUT«Unable to parse blockoid, couldn't find final '}' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
arnsholt masak: Point, point
masak rakudo: subset B of Any; sub classify($_) { when B {}; when B {} }
p6eval rakudo cb4521: ( no output )
masak confused
masak submits rakudobug
rakudo: subset Quad of Any; sub classify($_) { when Quad {}; when Quad {} }; say 'alive' 13:16
p6eval rakudo cb4521: OUTPUT«Unable to parse blockoid, couldn't find final '}' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
masak rakudo: subset B of Any; sub classify($_) { when B {}; when B {} }; say 'alive'
p6eval rakudo cb4521: OUTPUT«alive␤»
pausenclown_ subset Q of Any; sub classify($_) { when Q {}; when Q {} }; say 'alive' 13:18
masak rakudo: subset Q of Any; sub classify($_) { when Q {}; when Q {} }; say 'alive' 13:18
p6eval rakudo cb4521: OUTPUT«Missing block at line 11, near "; when Q {"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
masak .oO( things beginning with Q... ) 13:19
pausenclown_ what the fuck?
masak or, as they say in Latin, 'quod the fuck?'
arnsholt ^^
pausenclown_ rakudo: subset _Q of Any; sub classify($_) { when _Q {}; when _Q {} }; say 'alive'
p6eval rakudo cb4521: OUTPUT«alive␤»
pausenclown_ rakudo: subset q of Any; sub classify($_) { when q {}; when q {} }; say 'alive' 13:20
p6eval rakudo cb4521: OUTPUT«Missing block at line 11, near "; when q {"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pausenclown_ rakudo: say q(1)
p6eval rakudo cb4521: OUTPUT«Could not find sub &q␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
pausenclown_ rakudo: say q
p6eval rakudo cb4521: OUTPUT«Confused at line 11, near "say q"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
masak oh. 13:21
pausenclown_ what the heck is special about a Q?
masak we had a ticket like this the other day.
it's the quoting constructs.
the LTM should catch that, but it doesn't.
masak finds-and-merges
pausenclown_: thanks :)
pausenclown_ but there is no q construct anymore 13:22
masak there isn't?
rakudo: say q[oh hai!]
p6eval rakudo cb4521: OUTPUT«oh hai!␤»
pausenclown_ i feel like the dumb guy in the movie who has the idea without getting it. 13:23
muixirt rakudo: my $a=5; if $a>2 and $a<6 { say "YES"; }
p6eval rakudo cb4521: OUTPUT«YES␤»
muixirt rakudo: my $a=5; if $a<2 and $a>6 { say "YES"; }
p6eval rakudo cb4521: OUTPUT«Missing block at line 11, near "6 { say \"Y"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pausenclown_ rakudo: my $a=5; if 2 < $a <6 { say "YES"; }
p6eval rakudo cb4521: OUTPUT«YES␤»
snarkyboojum bonjour pausenclown
whoops - that was delayed somewhat :) 13:24
pausenclown_ ieeeek! french!!
muixirt why is the my second code snippet wrong?
pausenclown_ hides
snarkyboojum :)
arnsholt muixirt: The < > might get interpreted as a hash accessor
muixirt arnsholt: is that a bug or a feature? 13:25
snarkyboojum hi masak-san
pausenclown_ besides that, a number can never be smaller than 2 and bigger than 6 at the same time?
snarkyboojum er masak_san
colomon bug
snarkyboojum :)
arnsholt muixirt: I've no idea. But if colomon says bug, that's probably right =)
masak snarkyboojum: hi!
SmokeMachine: (neither one highlights here)
colomon I'm just going on instinct there, but I'm pretty sure it must be a bug.
masak er, meant snarkyboojum. 13:26
moritz_ rakudo: say ?( 6 > 1|7 < 2)
muixirt well why is that:
p6eval rakudo cb4521: OUTPUT«1␤»
moritz_ yes, we can!
muixirt pugs: my $a=5; if $a<2 and $a>6 { say "YES"; }
masak colomon: no, it's a hash indexing.
p6eval pugs: OUTPUT«*** ␤ Unexpected "6"␤ expecting term postfix, operator or bare or pointy block construct␤ at /tmp/0NmGRO3vHq line 1, column 25␤»
moritz_ muixirt: if there's no whitespace after $a, it parses the < as a postcircumfix
muixirt that makes me think that it is not a bug
colomon masak: I guess if I squint my eyes I can see that... 13:27
moritz_ std: my $a=5; if $a<2 and $a>6 { say "YES"; }
p6eval std 30383: OUTPUT«===SORRY!===␤Missing block at /tmp/iQe7Rvy7yS line 1:␤------> my $a=5; if $a<2 and $a>⏏6 { say "YES"; }␤ expecting any of:␤ POST␤ bracketed infix␤ infix or meta-infix␤ postfix␤ postfix_prefix_meta_operator␤Parse
..failed␤FAILED 00:01 108m␤»
colomon std: my $a=5; if $a<2 and $a>6 { say "YES"; }
p6eval std 30383: OUTPUT«===SORRY!===␤Missing block at /tmp/J5KU8nYkL1 line 1:␤------> my $a=5; if $a<2 and $a>⏏6 { say "YES"; }␤ expecting any of:␤ POST␤ bracketed infix␤ infix or meta-infix␤ postfix␤ postfix_prefix_meta_operator␤Parse
..failed␤FAILED 00:01 108m␤»
moritz_ so rule No. 1 of my not-yet-written style guide: always use whitespace around infix operators
colomon indeed
pausenclown_ std: my $a=5; if $a < 2 and $a > 6 { say "YES"; }
masak aye.
p6eval std 30383: OUTPUT«ok 00:01 108m␤»
colomon moritz_++
masak unless you're golfing and really know what you're doing. 13:28
pausenclown_ whitespace++ =)
muixirt moritz_: and there is no solution to that?
moritz_ muixirt: there is. Whitespace.
masak rakudo: say " "x 9-$_,"#"x$_*2-1 for 0..9,2 xx 3
p6eval rakudo cb4521: OUTPUT« ␤ #␤ ###␤ #####␤ #######␤ #########␤ ###########␤ #############␤ ###############␤#################␤ ###␤ ###␤ ###␤»
masak chopped the Christmas Tree into slices :) 13:29
moritz_ masak: ay, but in the case of golfing I remove the whitespaces as an extra pass in the end
masak *nod*
pausenclown_ rakudo: my $a=5; if $a ~~ <2|<6 { say "YES"; }
p6eval rakudo cb4521: OUTPUT«Confused at line 11, near "if $a ~~ <"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
pausenclown_ rakudo: my $a=1; if $a ~~ any({<2},{<6} { say "YES"; } 13:30
p6eval rakudo cb4521: OUTPUT«Unable to parse blockoid, couldn't find final '}' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
colomon afk # storytime
moritz_ #11960 So you threw in some random punctuation for no particular reason, and then you didn't get the result you expected. Hmmmm.
MJD++
pausenclown_ rakudo: my $a=1; if $a ~~ any({$_<2},{$_<6} { say "YES"; }
p6eval rakudo cb4521: OUTPUT«Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
muixirt moritz_: it's confusing this whitespace sensitivity (esp. with the not so helpful error messages that come along with these problems)
masak muixirt: it's a conscious tradeoff. 13:31
muixirt: and the error messages are getting steadily better.
moritz_ muixirt: yes. But it's good for educating you to use a sane code layout
pausenclown_ cant i make a juntion of codeblocks?
*junction
moritz_ pausenclown_: you can, but you need to use code blocks for that :-)
rakudo: say ?(-2 ~~ { $_+2 } | { $_-2}) 13:32
p6eval rakudo cb4521: OUTPUT«1␤»
moritz_ rakudo: say ?(-1 ~~ { $_+2 } | { $_-2})
p6eval rakudo cb4521: OUTPUT«1␤»
moritz_ rakudo: say ?(-2 ~~ { $_+2 } & { $_-2})
p6eval rakudo cb4521: OUTPUT«0␤»
pausenclown_ rakudo: my $a=1; if $a ~~ any({$_ < 2},{$_ < 6} { say "YES"; } 13:33
p6eval rakudo cb4521: OUTPUT«Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
pausenclown_ rakudo: my $a=1; if $a ~~ any({$_ < 2},{$_ < 6}) { say "YES"; }
p6eval rakudo cb4521: OUTPUT«YES␤»
pausenclown_ watch ur braces! =)
rakudo: my $a=3; if $a ~~ any({$_ < 2},{$_ < 6}) { say "YES"; }
p6eval rakudo cb4521: OUTPUT«YES␤» 13:34
pausenclown_ rakudo: my $a=3; if $a ~~ any({$_ < 2},{$_ > 6}) { say "YES"; }
p6eval rakudo cb4521: ( no output )
masak snarkyboojum: I still haven't paid back that technical debt with testing the Yapsi runtime...
snarkyboojum: but after that, it's time to take a bite out of immediate blocks :) 13:35
snarkyboojum masak: sounds fun
masak snarkyboojum: things like 'my $a = 42; { say $a }'
I think it will be.
snarkyboojum masak: ok
masak: so runtime.t needs filling out? 13:36
moritz_ wonders if Perl 6 ASTs needs to be really split up in two trees - one primarily sorted by scope, and one by phaser 13:37
masak snarkyboojum: yes. if you want to help, we might divide the labour as one making up tests, and the other one adding them. 13:38
snarkyboojum masak: always interested and willing to help :) will have a look 13:39
masak moritz_: the thing about phasers is that they aren't all there at parse time.
moritz_ masak: nothing is there until it's parsed... that's kinda obvious, no? 13:40
masak moritz_: sure, but not what I meant. 13:41
I meant that phasers can be supplied at runtime as well.
or somewhere in between.
I've also been considering a 'static model' for the phasers, and more or less concluded that they will only get you so far. 13:42
moritz_ they can? but only in an eval, no?
masak or in a macro.
moritz_ well, a macro is expanded at BEGIN time
so for all following phasers it's really irrelevant if it was supplied by a phaser or not 13:43
masak is BEGIN time necessarily over before the parsing is over?
moritz_ yes 13:43
it's as soon as the BEGIN block/statement is parsed 13:44
masak it only says 'as soon as possible'...
that might be a bit later sometimes.
moritz_ that's as soon as possible :-)
if you talk about "a bit later", you already know it's not "as soon as possible" :-)
masak .oO( whoa, named enums are like hashes, subtypes and subs all rolled into one... )
moritz_ I mean you couldn't change the grammar very effectfully if BEGIN time was after the parse of the whole file 13:45
masak moritz_: well, we're all a bit late sometimes for various reasons...
masak std: BEGIN { foo() }; sub foo() { say "OH HAI" } 13:46
p6eval std 30383: OUTPUT«ok 00:01 110m␤»
masak moritz_: has the 'foo' been parsed when the BEGIN block runs?
moritz_ masak: sure 13:46
masak the sub, I mean.
then the parse is all done.
moritz_ it first parses up to to the last closing }
masak: yes, but that's a special case... there could be text after the BEGIN block 13:47
masak so the BEGIN block runs after the parse, not "as soon as possible".
moritz_ it runs after the parse *of the block*
not after the parse of the compilation unit
masak oh, I see. 13:48
moritz_ BEGIN { A }; B; BEGIN { C }; D
first parses BEGIN { A };
runs A at compile time
then parses B;
then parses BEGIN {C }
runs C
parses D
and if A or C contain more BEGIN blocks, apply recursively 13:49
Juerd It doesn't really have anything to do with beginning, does it?
masak waitwait.
moritz_ Juerd: not anymore 13:50
masak moritz_: if it runs A before it parses B, how can it work in my example?
Juerd NOW { ... } ;)
masak moritz_: I mean, the sub 'foo' isn't declared until after the BEGIN block.
moritz_ masak: oh sorry, I mispared your example
masak: I thought the sub foo () { } was inside the BEGIN block
then it won't work
masak but STD.pm6 doesn't detect it? 13:51
moritz_ seems like
masak: perl 5 also dies
perl -e 'BEGIN { foo() }; sub foo { }'
Undefined subroutine &main::foo called at -e line 1.
BEGIN failed--compilation aborted at -e line 1.
moritz_ sorry for the confusion 13:52
masak fair enuf. 13:52
moritz_ rakudo: { foo() }; sub foo { say 3 } 13:52
p6eval rakudo cb4521: OUTPUT«3␤»
masak Rakudo works too well in this case. :)
moritz_ that only works because &foo is entered into the symbol table of the surrounding compilation unit
rakudo: BEGIN { foo() }; sub foo { say 3 }
p6eval rakudo cb4521: OUTPUT«Could not find sub &foo␤current instr.: '_block11' pc 0 (EVAL_1:9)␤»
moritz_ rakudo++
masak aha. 13:53
right; the first one had no BEGIN in it.
moritz_ note the missing "non-existent" part in the error message :-)
masak \o/ 13:54
someone++
masak Juerd: well, NOW is not such a good name, because when 'now' is depends on when now is :) 14:00
moritz_ ASAP { ... }
masak cute :) 14:01
and END { ... } would be ALAP { ... } :)
Juerd I like ASAP :) 14:02
huf for a second there i thought the channel has switched to hungarian... 14:03
Juerd masak: That would imply that all ALAPs run in parallel :)
huf dont just throw around "words", you dont know what language you'll hit
masak Google Translate says 'alap' means 'base' 14:04
huf yep
masak .oO( all your alap... )
moritz_ all your ALAP are...
huf :DDDD
Juerd I was looking at my irc window, thought "all your alap" and then saw it there, as if the computer had read my mind and typed it for me, only your nicknames were in front of the lines, not mine :) 14:05
masak it was too obvious, it seems :)
Juerd Obvious is good. 14:06
And clichés exist because they're true ;)
moritz_ principle of least surprise :-)
masak :)
moritz_ but to be funny, you have to surprise people
Juerd Not necessarily
Some forms of humor depend on the audience knowing exactly what's going to happen 14:07
masak you could meta-surprise them by not surprising them :)
Juerd For example, if a certain culture finds something funny unconditionally.
diakopter masak: a higher blah/comment ratio can be seen in the commits at sprixel/perlesque at csmeta.org ;) 14:10
masak oh noes! :)
here? code.google.com/p/csmeta/updates/list 14:11
I see no bla-s at all...
moritz_ but emty commit messages 14:11
masak those aren't bla-s :) 14:12
the point is that actually typing 'bla bla bla' in the commit message is somehow *less* informative than putting in nothing :)
it conveys 'I know this is a commit message, but I can't be bothered". 14:13
diakopter but blank ones...?
masak those might mean any number of things. 14:14
'this commit is unworthy of a comment'
'I don't know what to type here'
'I somehow didn't realize that it was possible to leave a commit message'
moritz_ 'Others don't read my code anyway, and I myself don't really care all that much'
pausenclown_ how do i write this shorter? 14:15
masak 'I have better things to do with my life than describe what changes I make. it was hard to write, it should be hard to figure out!'
pausenclown_ self.books[*-1]{$context.name} ??
self.books[*-1]{$context.name}.push( $text.data ) !!
self.books[*-1]{$context.name} = [ $text.data ] ;
masak pausenclown_: with proper autovivification, it should just be the second line.
pausenclown_: for now, write a sub that does it for you.
taking a hash and a key as arguments. 14:16
moritz_ given self.books[*-1] { .{$context.name} //= []; .push: $text.data }
masak er, a hash, a key, and an array element.
or what moritz_ wrote :)
moritz_ erm, forgot one .{$context.name}
pausenclown_ masak: that was my expectation but it failed in rakudo. 14:18
masak what failed? the encapsulation?
pausenclown_ Method 'push' not found for invocant of class 'Proxy' 14:20
masak please nopaste the whole attempt.
pausenclown_ rakudo: class Foo { has @.a is rw; method foo { self.a[0].push('foo'); } }; my $a = Foo.new; $a.foo; say $a.a 14:27
p6eval rakudo cb4521: OUTPUT«Method 'push' not found for invocant of class 'Proxy'␤current instr.: 'perl6;Foo;foo' pc 451 (EVAL_1:181)␤»
masak rightright, but you still need to do the trick of manually supplying the inner array *within* the method. 14:28
the thing with the encapsulation into a sub is that you only have to go through that once.
masak rakudo: sub create-or-push(%h is rw, $key, $elem) { %h{$key} ?? %h{$key}.push($elem) !! (%h{$key} = [$elem]) }; my %hash; create-or-push(%hash, "foo", $_) for 42, "OH HAI"; say %hash<foo>.perl 14:32
p6eval rakudo cb4521: OUTPUT«[42, "OH HAI"]␤»
masak something like that. 14:33
pausenclown_ rakudo: class Foo { has @.a is rw; method foo { self.a = self.a ?? [ |self.a, 'foo'] !! ['foo']; } }; my $a = Foo.new; $a.foo; say $a.a
p6eval rakudo cb4521: OUTPUT«foo␤»
pausenclown_ timtowdy =)
masak I suspect your way will give you arbitrary deep nestings of arrays. 14:34
s/arbitrary/arbitrarily/
pausenclown_ you mean | flattens nested?
rakudo: say ( | [1,[2,3]]).perl 14:35
p6eval rakudo cb4521: OUTPUT«Capture.new()␤»
pausenclown_ rakudo: say | [1,[2,3]].perl
p6eval rakudo cb4521: OUTPUT«[1, [2, 3]]␤»
masak no, I mean assigning to a variable with the @ sigil already creates one layer. [] creates another. 14:36
pausenclown_ yeah, right. 14:38
masak since you're taking self.a and assigning [] to it every time, you'll get a new array layer every time.
pausenclown_ rakudo: class Foo { has @.a is rw; method foo($x) { self.a = self.a ?? (self.a, $x) !! $x; } }; my $a = Foo.new; $a.foo(1); $a.foo(2); $a.foo(3); say $a.a.perl; 14:41
p6eval rakudo cb4521: OUTPUT«[1, 2, 3]␤»
pausenclown_ thats it
pausenclown_ i indeed forgot about the influence of the @ 14:41
masak it's insidiously easy to get double layers in different ways. 14:42
rakudo: class A { has @.b }; say A.new(:b[1,2,3]).b.perl
p6eval rakudo cb4521: OUTPUT«[[1, 2, 3]]␤»
masak that's a bug nowadays, due mostly to my extensive complaining. 14:43
masak rakudo: .say for map -> $a, $b { "$a, $b" }, 1, 2, 3, 4 14:57
p6eval rakudo cb4521: OUTPUT«Too many positional parameters passed; got 5 but expected 2␤current instr.: 'map' pc 341553 (src/gen/core.pir:18789)␤»
masak :(
colomon rakudo: map -> $a, $b { "$a, $b" }, 1, 2, 3, 4 14:59
p6eval rakudo cb4521: OUTPUT«Too many positional parameters passed; got 5 but expected 2␤current instr.: 'map' pc 341553 (src/gen/core.pir:18789)␤»
colomon rakudo: map -> $a, $b { "$a, $b" }, (1, 2, 3, 4)
p6eval rakudo cb4521: ( no output )
colomon shouldn't let masak sidetrack him like this... 15:00
masak rakudo: for 1, 2, 3 X <a b c> -> $a, $b { say $a, $b } 15:03
p6eval rakudo cb4521: OUTPUT«1 a1 b␤1 c2 a␤2 b2 c␤3 a3 b␤StopIteration␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
masak wow, it's actually brokeneder than before the merge.
colomon before what merge? 15:05
moritz_ ng
masak the ng one.
now it's wrong in two ways. 15:06
masak before it was only wrong in one way. 15:06
colomon ah. 15:07
I've fixed .say for map -> $a, $b { "$a, $b" }, 1, 2, 3, 4 locally, but then map.t blows up. 15:08
colomon No applicable candidates found to dispatch to for 'map'. Available candidates are: 15:09
:(Mu : &block;; *%_)
what's with the Mu? I keep on getting that no matter what (this one is referring to Any.map, for instance...) 15:10
moritz_ invocant marker?
moritz_ is the sub form of map defined as 'is export' of a method? 15:13
colomon proto sub map(&mapper, *@values) { @values.map(&mapper); } 15:14
that's my hacked version
proto sub map(&mapper, @values) { @values.map(&mapper); }
is the normal one.
moritz_ doesn't the *@stuff has still problems?
colomon I don't know.
I'm sure thats why it wasn't *@stuff two months ago.
moritz_ I tried to change grep to use a slurpy about two weeks ago, and it blew up 15:15
colomon but things are getting fixed all the time, and some of the protos in Any-list.pm are already using *@a
At any rate, the Mu thing is orthogonal to that, I think. 15:16
moritz_ rakudo: my @a = 1, 2; say @.PARROT 15:17
p6eval rakudo cb4521: OUTPUT«Lexical 'self' not found␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
moritz_ rakudo: my @a = 1, 2; say @a.PARROT
p6eval rakudo cb4521: OUTPUT«Array␤»
moritz_ rakudo: sub f(*@a) { say @a.PARROT }; f(1, 2) 15:18
p6eval rakudo cb4521: OUTPUT«Array␤»
colomon didn't know .PARROT worked.
moritz_ colomon: changing the @values to *@values in grep works, except that it then can't be called by name anymore 15:22
colomon ah
moritz_ which is less sever than not working on single items, IMHO
rakudo: proto sub a(*@val) { say @val.perl }; a(:val<a b c>) 15:23
p6eval rakudo cb4521: OUTPUT«Unexpected named parameter 'val' passed␤current instr.: 'a' pc 223 (EVAL_1:83)␤»
moritz_ rakudo: proto sub a(@val) { say @val.perl }; a(:val<a b c>)
p6eval rakudo cb4521: OUTPUT«("a", "b", "c")␤»
colomon just undid my patch locally so I can get to $work using my Rakudo build. :)
moritz_ submits rakudobug
colomon moritz_++
moritz_ rakudo: proto sub a(@val) { say @val.perl }; say &a.signature.perl 15:25
p6eval rakudo cb4521: OUTPUT«get_attr_str() not implemented in class 'Perl6MultiSub'␤current instr.: 'perl6;Code;signature' pc 13230 (src/builtins/Str.pir:115)␤»
moritz_ rakudo: sub a(*@val) { say @val.perl }; a(:val<a b c>)
rakudo: sub a(*@val) { say @val.perl }; a(:val<a b c>))
p6eval rakudo cb4521: OUTPUT«Unexpected named parameter 'val' passed␤current instr.: 'a' pc 194 (EVAL_1:78)␤»
p6eval rakudo cb4521: OUTPUT«Confused at line 11, near ")"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤» 15:25
moritz_ rakudo: sub a(*@val) { say @val.perl }; say &a.signature.perl
p6eval rakudo cb4521: OUTPUT«:(*@val)␤» 15:26
moritz_ rakudo: sub a(@val) { say @val.perl }; say &a.signature.perl
p6eval rakudo cb4521: OUTPUT«:(@val)␤»
moritz_ rakudo: sub a(@val) { say @val.perl }; a(:val<foo bar>)
p6eval rakudo cb4521: OUTPUT«("foo", "bar")␤»
pugssvn r30384 | moritz++ | [t/spec] change fudging in grep.t for rakudo 15:29
dalek kudo: 3b5a790 | moritz++ | src/core/Any-list.pm:
make the grep() arguments slurpy
 15:31
masak huh :)
moritz_ masak: but don't worry, I've replaced it with a new one :-) 15:32
RT #74344 15:33
masak oh phew :)
oh, that was the bug.
moritz_ wait
moritz_ just confused bug numbers
the one in the commit message is wrong :( 15:34
masak hm, yes.
moritz_ nope 15:35
7*34*44 vs 7*43*44
masak :) 15:36
rakudo: say <a b>.pick(*).perl; say (<a b>.pick(*) X <a b>.pick(*)).perl 15:39
p6eval rakudo cb4521: OUTPUT«("a", "b")␤()␤»
masak submits rakudobug
for some reason, pick(*) and X don't interoperate well.
moritz_ rakudo: say <a b>.pick(*).PARROT
p6eval rakudo cb4521: OUTPUT«GatherIterator␤»
masak GatherIterator. my nemesis. 15:40
colomon Actually, that's probably a gather / take bug
masak oh no. 15:40
colomon rakudo: say (<a b>.pick(*) X <a b>).perl
p6eval rakudo cb4521: OUTPUT«(("a", "a"), ("a", "b"), ("b", "a"), ("b", "b"))␤»
colomon rakudo: say (<a b>.pick(*) X <a b>).perl
p6eval rakudo cb4521: OUTPUT«(("a", "a"), ("a", "b"), ("b", "a"), ("b", "b"))␤»
masak colomon++
masak I'm afraid I have to reach the conclusion that Rakudo master isn't ready for my poker example just yet. 15:41
colomon Any time you use the same code twice and get a GatherIterator back both times, you can expect trouble.
masak the workarounds are piling up to infinity.
colomon: so it's not really a new bug, then?
colomon nope.
masak doesn't submit 15:42
hm, but I can work around it, by doing one at a time...
moritz_ I'm sure you can add .eager to them 15:43
colomon rakudo: say (<a b>.pick(*) X <a b>.pick(*).Seq).perl
p6eval rakudo cb4521: OUTPUT«()␤»
moritz_ rakudo: say (<a b>.pick(*).eager X <a b>.pick(*).eager).perl
p6eval rakudo cb4521: OUTPUT«(("b", "a"), ("b", "b"), ("a", "a"), ("a", "b"))␤»
masak assigning one of them to an array worked !
colomon rakudo: say (<a b>.pick(*).Seq X <a b>.pick(*)).perl
p6eval rakudo cb4521: OUTPUT«(("b", "b"), ("b", "a"), ("a", "b"), ("a", "a"))␤»
masak oh, eager works too. nice.
colomon masak: assigning one to an array only works because arrays aren't properly lazy at the moment. :( 15:44
masak I thought array assignment was supposed to be mostly eager.
colomon masak: you might be right about that, mostly eager is my nemesis. :)
masak we all have our nemeses :) 15:45
rakudo: subset Foo of Array where { .elems == 5 }; sub bar(Foo $_) {}; bar([1,2,3]) 15:47
p6eval rakudo cb4521: OUTPUT«Nominal type check failed for parameter '$_'; expected Array but got Array instead␤current instr.: 'bar' pc 468 (EVAL_1:179)␤»
masak submits rakudobug
it's not the nominal typecheck that fails, it's... the other one.
the one with the where clause, whatever it's called.
moritz_ the nonominal type check :-) 15:48
masak really? I thought "nominal" had to do with the name part, i.e. Array in this case. 15:49
and that there was another part of the type check (possibly) called something else.
moritz_ nonominal = non-nominal :-)
arnsholt masak: Has such trifling details ever stopped Perl 6? =)
moritz_ just made that up
masak oh, ah. 15:50
arnsholt Heh. I bought it =)
masak wishes for a om-nominal type check :)
s/a/an/
moritz_ procedural type check? 15:51
TimToady jnthn++ started calling it "nominal" because it's part of the longname of a multi candidate
masak moritz_: that sounds vaguely right. 15:52
TimToady and because it's pretty much limited to a named type
moritz_ at lest S05 distinguishes "procedural" and "declarative" elements
so we could steel that "procedural" for this case 15:53
TimToady you run the nominal check by (notionally) checking the name directly, not by running constraint code
moritz_ rakudo: sub &foo { }; say &foo.name
p6eval rakudo 3b5a79: OUTPUT«Confused at line 11, near "sub &foo {"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
moritz_ rakudo: sub foo { }; say &foo.name 15:54
p6eval rakudo 3b5a79: OUTPUT«foo␤»
masak rakudo: class G {}; subset A of Array where { all(|$_) ~~ G }; say [G.new, G.new, G.new] ~~ A
p6eval rakudo 3b5a79: OUTPUT«0␤»
masak :( 15:55
masak seems to be running headlong into Rakudo today
rakudo: class G {}; my @g = [G.new, G.new, G.new]; say all(|@g) ~~ G 15:57
p6eval rakudo 3b5a79: OUTPUT«0␤»
masak oh.
moritz_ @g contains an Array 15:58
masak rakudo: class G {}; my @g = G.new, G.new, G.new; say all(|@g) ~~ G
right.
p6eval rakudo 3b5a79: OUTPUT«0␤»
moritz_ rakudo: class G {}; my @g = G.new, G.new, G.new; say all(@g) ~~ G
p6eval rakudo 3b5a79: OUTPUT«0␤»
moritz_ rakudo: class G { }; say all(G.new) ~~ G
p6eval rakudo 3b5a79: OUTPUT«0␤»
moritz_ rakudo: class G { }; say G.new ~~ G 15:59
p6eval rakudo 3b5a79: OUTPUT«1␤»
masak submits rakudobug
moritz_ rakudo: class G { }; say ?(all( G.new ) ~~ G)
p6eval rakudo 3b5a79: OUTPUT«0␤»
masak rakudo: say all(42) ~~ Int
p6eval rakudo 3b5a79: OUTPUT«0␤»
masak maybe junctional smartmatching simply isn't in place yet.
moritz_ rakudo: say ('foo'&'bar').uc 16:00
p6eval rakudo 3b5a79: OUTPUT«all("FOO", "BAR")␤»
moritz_ rakudo: say ('foo'&'bar') ~~ .uc
p6eval rakudo 3b5a79: OUTPUT«0␤»
TimToady maybe the .ACCEPTS has a Mu arg?
moritz_ I guess the appropriate ACCEPTS method is written in PIR still 16:01
and autothreading over routines written in PIR never worked
masak works around 16:02
dalek ok: 0732332 | moritz++ | src/subs-n-sigs.pod:
[subs] interpolating arguments and slurpy parameters
TimToady rakudo: class G { }; say G.ACCEPTS(any(Int,G.new)) 16:03
p6eval rakudo 3b5a79: OUTPUT«0␤»
ruoso rakudo: class G { }; say G.can('ACCEPTS').signature.perl 16:05
p6eval rakudo 3b5a79: OUTPUT«Method 'signature' not found for invocant of class 'P6Invocation'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
ruoso rakudo: class G { }; say G.get_method('ACCEPTS').signature.perl 16:06
p6eval rakudo 3b5a79: OUTPUT«Method 'get_method' not found for invocant of class ''␤current instr.: '_block14' pc 29 (EVAL_1:0)␤»
ruoso hmm
masak rakudo: subset Even of Int where { $_ !% 2 }; subset Fivey of Int where { $_ !% 5 }; subset Tennis of Int where Even & Fivey; say 40 ~~ Tennis
p6eval rakudo 3b5a79: OUTPUT«1␤»
masak \o/
ruoso: .can is totally b0rken as I understand it. don't use. 16:07
colomon \o/
moritz_ masak: you do know that 15 is a valid Tennis score? :-) 16:08
masak moritz_: yes, I was just giving the subset frivolous names :)
moritz_: at first I went for Tennish 16:09
moritz_ :-)
Teeny :-)
ruoso rakudo: subset Even of int where { $_ !% 2 }; subset Fivey of Int where { $_ !% 5 }; subset SmallerThen 41 of Int where { $_ < 41 }; subset Tennis of Int where Even & Fivey & SmallerThan41; say 40 ~~ Tennis; say 41 ~~ Tennis; 16:09
p6eval rakudo 3b5a79: OUTPUT«Confused at line 11, near "subset Eve"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
ruoso rakudo: subset Even of int where { $_ !% 2 }; subset Fivey of Int where { $_ !% 5 }; subset SmallerThen41 of Int where { $_ < 41 }; subset Tennis of Int where Even & Fivey & SmallerThan41; say 40 ~~ Tennis; say 41 ~~ Tennis;
p6eval rakudo 3b5a79: OUTPUT«Confused at line 11, near "subset Eve"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
ruoso bah
ruoso 's brain out of sync today 16:10
TimToady std: subset Even of int where { $_ !% 2 }; subset Fivey of Int where { $_ !% 5 }; subset SmallerThen41 of Int where { $_ < 41 }; subset Tennis of Int where Even & Fivey & SmallerThan41; say 40 ~~ Tennis; say 41 ~~ Tennis; 16:14
p6eval std 30384: OUTPUT«===SORRY!===␤Undeclared name:␤ 'SmallerThan41' used at line 1␤Check failed␤FAILED 00:01 110m␤»
TimToady std: subset Even of int where { $_ !% 2 }; subset Fivey of Int where { $_ !% 5 }; subset SmallerThan41 of Int where { $_ < 41 }; subset Tennis of Int where Even & Fivey & SmallerThan41; say 40 ~~ Tennis; say 41 ~~ Tennis; 16:15
p6eval std 30384: OUTPUT«ok 00:01 108m␤»
masak rakudo: subset A of Array; subset B of A; subset C of A; subset D of A where B & C; say [] ~~ D 16:16
p6eval rakudo 3b5a79: OUTPUT«Type objects are abstract and have no attributes, but you tried to access ␤current instr.: 'perl6;SeqIter;get' pc 14351 (src/builtins/Routine.pir:126)␤»
masak submits rakudobug
rakudo: subset A of Array; subset B of A; subset C of Any; subset D of A where B & C; say [] ~~ D 16:18
p6eval rakudo 3b5a79: OUTPUT«Type objects are abstract and have no attributes, but you tried to access ␤current instr.: 'perl6;SeqIter;get' pc 14351 (src/builtins/Routine.pir:126)␤»
masak it appears to be a combination of doing the nominal A check multiple times, and the &
rakudo: subset A of Array; subset B of Any; subset C of Any; subset D of A where B & C; say [] ~~ D 16:20
p6eval rakudo 3b5a79: OUTPUT«1␤»
ash_ so... was perl 5.12 release yesterday? 16:22
takadonet yes
moritz_ seems like
ash_ neat, perl.org still says 5.10.1 16:23
tis why i ask
CokeBot9000 search.cpan.org/dist/perl-5.12.0/ 16:24
ash_ thanks, CokeBot9000++ 16:24
CokeBot9000 np. 16:26
ash_ pluggable keywords? could you define new operators with that? 16:27
masak obra++ 16:28
masak I finally got the pokerhand script working: gist.github.com/364800 16:33
I'm exhausted :)
will attempt to get it into the book later today. 16:34
(I need to annotate it with all the workarounds I put in, so we can remove them as the bugs get fixed)
moritz_ ash_: keyword != operator 16:35
CokeBot9000 masak: ThreeOfAKind:Trips::FourOfAKind:Quad
I like Kwad. Nice workaround. =-) 16:36
ash_ well, i figured you could associate a symbol like ! or something, since it says it accepts any ascii characters as the keyword, but yeah, your right, that probably wouldn't work out nicely
[particle] Kwad? 16:37
CokeBot9000 masak: would you like a patch for the last XXX ?
[particle]: Quad triggers a rakudo bug (parsed like Q:uad, or something)
masak++ 16:38
[particle] uggh
masak CokeBot9000: yes, please!
theater & 16:39
CokeBot9000 latest rakudo doesn't build for me: 16:44
Lexical '$test' not found
current instr.: 'perl6;Perl6;Grammar;is_name' pc 41059 (src/gen/perl6-grammar.pi
(that's after a git pull and a realclean)
moritz_ looks 16:45
moritz_ CokeBot9000: I have a local patch here, but that shouldn't change much... builds here without problems 16:53
CokeBot9000: do you have a new parrot?
CokeBot9000: iirc somebody broke some lexicals handling stuff in parrot today, and I'm still using an older parrot
CokeBot9000 ah, yes, t/op/lexicals.t is failing for me in trunk. that's probably it. 16:54
dalek kudo: fb38dbf | quester++ | src/core/ (3 files):
Fix Parcel.sort and add Hash.sort.

  cloud.github.com/downloads/perl6/bo...010-04.pdf
See also: rt.perl.org/rt3/Ticket/Display.html?id=74334 Signed-off-by: Moritz Lenz [email@hidden.address]		 16:58
CokeBot9000 moritz_: fixed. 17:31
CokeBot9000 runs masak's script. (hurm. that's not quite right.) 17:32
moritz_ CokeBot9000++
sjohnson 3hi moritz_ 17:32
oops 17:33
CokeBot9000 nopaste? 17:52
moritz_ see /topic
lisppaste3 CokeBot9000 pasted "Minor updates for masak's poker script; deal one at a time, s/Any/PokerHand/, fix issue with where clause in FullHouse. " at paste.lisp.org/display/97730 17:55
CokeBot9000 (without the update to fullhouse, /everything/ was a full house.
moritz_ phenny: tell masak about paste.lisp.org/display/97730 by CokeBot9000++ 17:57
phenny moritz_: I'll pass that on when masak is around.
CokeBot9000 moritz++ 18:05
moritz_ dammit, why can't network routing be easy? 18:06
diakopter job security? 18:07
the eternal struggle. job security vs innovation.
moritz_ I have a simple setup where I want my laptop to act as a router between my desktop and the router provided by the ISP 18:08
and so far I've tried three different configurations, and after some debugging I understood why each of them didn't work
but so far I'm refusing to do any internal NAT routing or masquerading 18:09
bkeeler So #rs is soon, yes? 18:19
arnsholt moritz_: Why do you need a router behind the other router? 18:23
moritz_ arnsholt: because I have no direct connection between router and PC 18:28
CokeBot9000 I heard you like routing...
moritz_ bkeeler: #rs is in 2 minutes, but due to a real life emergency I have to run off in a few minutes :( 18:29
arnsholt moritz_: You could try to get your laptop to bridge, rather than route? 18:32
Might be simpler
[particle] moritz_: do you know about www.vyatta.org/ ?
moritz_ has to run, will look at it later 18:34
mberends if anyone has anything to discuss in #rakudosketch, please join in the next few minutes 18:40
mberends moritz_: is your emergency under control now? 19:19
moritz_ mberends: yes
ingy I have a theoretical question about versioning... 19:26
more or less theoretical
I am trying to figure out a versioning scheme for C'Dent
and realized that there is a spec version and an implementation version 19:27
at least
maybe a standard lib spec version
what does perl 6 do?
WWP6D? 19:28
mberends there is a major section called Versioning in S11: perlcabal.org/syn/S11.html#Versioning 19:29
and a short term plan: github.com/rakudo/rakudo/blob/maste...oposal.pod 19:30
colomon rakudo: my @a = 1...10; say @a[*-1] 19:39
p6eval rakudo fb38db: OUTPUT«10␤»
colomon word.
lisppaste3 cognominal pasted "error using proto token in rakudo-nqp-rx" at paste.lisp.org/display/97741 20:11
moritz_ cognominal: proto *token* test <...> 20:13
moritz_ proto test declares a proto sub test 20:15
cognominal thx 20:16
I guess this the wrong default in a grammar. 20:17
moritz_ it is; but changing the default in different kind of scopes would be ETOOMUCHMAGIC 20:18
cognominal now suppose I want a blank as sym in proto token, what would be the synta for that? 20:19
Is ETOOMUCHMAGIC possible in a Perl interpreter?
moritz_ :sym< > 20:20
at least I hope
it's taken as a literal, not as a regex
or you just use ' ' instead of <sym> in the body of the rule
cognominal I am not sure, here blank has a special meaning token postfix:sym<[ ]> { 20:26
colomon rakudo: my @a = 1...3; for @a, 42 { .say } 20:30
p6eval rakudo fb38db: OUTPUT«1␤2␤3␤42␤»
colomon rakudo: my @a = 1...3; (@a, 42).map( { .say } ).eager
p6eval rakudo fb38db: OUTPUT«1␤2␤3␤42␤»
spinclad :sym(' ') might work, i expect :sym< > to trim whitespace to nothing 20:34
:sym<[ ]> means :sym('[', ']'), as usual, iirc 20:35
colomon rakudo: my @a = 1...3; (@a, 42).kv.eager.perl.say 20:37
p6eval rakudo fb38db: OUTPUT«(0, 1, 1, 2, 2, 3, 3, 42)␤»
colomon rakudo: my @a = 1...3; (@a, 42).pairs.eager.perl.say 20:38
p6eval rakudo fb38db: OUTPUT«(0 => 1, 1 => 2, 2 => 3, 3 => 42)␤»
colomon rakudo: my @a = 1...3; (@a, 42).pairs.min({ .value }).perl.say 20:40
p6eval rakudo fb38db: OUTPUT«Too many positional parameters passed; got 2 but expected between 0 and 1␤current instr.: '_block57' pc 413 (EVAL_1:147)␤»
colomon rakudo: my @a = 1...3; (@a, 42).pairs.min({ $^a.value <=> $^b.value }).perl.say 20:41
p6eval rakudo fb38db: OUTPUT«0 => 1␤»
solu Hey, I have a couple questions about importing modules into a Perl 6 script. 20:51
Can you import a Perl 5 module into Perl 6?
colomon not yet 20:52
solu Okay, thanks.
mberends it's being worked on over here: github.com/jnthn/blizkost
solu How do you change the @INC paths? In Perl 5, it was: use lib "my new directory"; is there a Perl 6 version of this? 20:53
mberends Perl 6 has @*INC for the same purpose 20:54
solu so it would be: push @*INC, "new directory"?
mberends yes, inside a BEGIN { ... } block 20:55
solu okay, thanks!
mberends if you put your directories into PERL6LIB before running your program, that also adds them to @*INC
solu hmm, still getting an error. 21:00
Can't find ./Module in @*INC in Main (src\gen_setting.pm:445) 21:01
I pushed the right directory, that's not the problem, I think. I'm using "use Module;" to import it
any idea what's wrong? 21:03
snarkyboojum rakudo: subset Quad of Any; sub classify($_) { when Quad {}; when Quad {} } 22:02
p6eval rakudo fb38db: OUTPUT«Unable to parse blockoid, couldn't find final '}' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
snarkyboojum I like that's related to the recent q-bug :)
rakudo: subset uad of Any; sub classify($_) { when uad {}; when uad {} }
p6eval rakudo fb38db: ( no output )
snarkyboojum rakudo: subset Q of Any; sub classify($_) { when Q {}; when Q {} } 22:03
p6eval rakudo fb38db: OUTPUT«Missing block at line 11, near "; when Q {"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
snarkyboojum rakudo: subset q of Any; sub classify($_) { when q {}; when q {} }
p6eval rakudo fb38db: OUTPUT«Missing block at line 11, near "; when q {"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
snarkyboojum rakudo: subset Qa of Any; sub classify($_) { when Qa {}; when Qa {} } 22:03
p6eval rakudo fb38db: OUTPUT«Unable to parse blockoid, couldn't find final '}' at line 11␤current instr.: 'perl6;Regex;Cursor;FAILGOAL' pc 1664 (ext/nqp-rx/src/stage0/Regex-s0.pir:907)␤»
snarkyboojum rakudo: subset a of Any; sub classify($_) { when a {}; when a {} }
p6eval rakudo fb38db: ( no output )
snarkyboojum :)
snarkyboojum rakudo: subset q of Any; when q {}; 22:18
p6eval rakudo fb38db: OUTPUT«Missing block at line 11, near ";"␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
snarkyboojum rakudo: subset q of Any; when /q/ {}; 22:19
p6eval rakudo fb38db: ( no output )
snarkyboojum rakudo; subset q of Any; when q(test) {};
rakudo: subset q of Any; when q/test/ {};
p6eval rakudo fb38db: ( no output )
snarkyboojum rakudo: when q {} 22:21
p6eval rakudo fb38db: OUTPUT«Missing block at line 11, near ""␤current instr.: 'perl6;HLL;Grammar;panic' pc 500 (ext/nqp-rx/src/stage0/HLL-s0.pir:328)␤»
oskie i wonder if rakudo will be debianized prior to or at release 22:28
ooh, it's in experimental
pmurias oskie: rakudo has loads releases 22:29
* loads of
pmurias would an erlang like concurrency model make sense in Perl 6? 22:37
CokeBot9000 blogs.activestate.com/2010/04/perl-5-is-alive/ - "Development of Perl 6 was announced in Summer 2000. Attempts to change Perl 5 in any major way were forthwith redirected to the Perl 6 project, as Perl 5 was now supposed to go into maintenance mode" ... really? 23:12
Juerd Somewhat. 23:13
Juerd Perl 6 turned out to become a new language instead :) 23:14
diakopter I think CokeBot9000 asks "really?" about the "was supposed to go into maintenance mode" portion
Juerd That view did exist. 23:15
Nobody knows how many people actually believed it :)
diakopter ah
Juerd I have no idea if this was ever an official thing, by the way. 23:16
SmokeDroid rakudo, @a = 1, 1 ... {$^a + $^b}; @a[3].say 23:38
m6locks you need to type rakudo: to evaluate 23:39
(i guess) 23:40
rakudo: @a = 1, 1 ... {$^a + $^b}; @a[3].say
p6eval rakudo fb38db: OUTPUT«Symbol '@a' not predeclared in <anonymous>␤current instr.: 'perl6;PCT;HLLCompiler;panic' pc 152 (compilers/pct/src/PCT/HLLCompiler.pir:108)␤»
quester_ rakudo: <1 3 4 2>.sort({-$^a}).say 23:43
p6eval rakudo fb38db: OUTPUT«4321␤»
quester_ :) 23:43
diakopter rakudo: my @a = 1, 1 ... {$^a + $^b}; @a[3].say 23:45
p6eval rakudo fb38db: ( no output )
lue hello 23:58

Powered by Cro and the Raku® Programming Language.

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

Thank you!