»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
lue hello world o/ 00:01
sorear lue! 00:04
lue hi sorear o/
dalek p/toqast: 407a1cf | pmichaud++ | src/HLL/Compiler.pm:
Update stagestats to use a bitmask instead of levels.

   0x04 = force gc sweep before displaying stats
   0x08 = pause between stages		 00:57
pmichaud arggh, pushed to the wrong branch. 01:14
I'll cherry-pick it into master.
dalek p: a4165a9 | pmichaud++ | src/HLL/Compiler.pm:
Update stagestats to use a bitmask instead of levels.

   0x04 = force gc sweep before displaying stats
   0x08 = pause between stages		 01:16
Lard Hi guys, Is there something similar to moose type coercions in perl6? 02:09
I can't find any docs on it... 02:10
"is also" is mentioned here gist.github.com/247656 but I can't see that documented anywhere 02:11
sorear Lard: no. 02:17
the closest we have are coerced types, like has Int() $.foo 02:18
you can initialize $.foo with any value, and the .Int method will be called
but you can't do custom coercers like Moose :(
also I'm not sure that anyone implements coerced types yet.
what you can do *today* is write a custom BUILD method which initializes the attributes 02:19
Lard sorear: ah ok thanks... I'll see if it that works now...
sorear submethod BUILD(:$foo) { $!foo = $foo.Int }
Lard really I couldn't get BUILD to work....
I do submethod BUILD ( :$foo ) {$!bar = $foo;} 02:20
having declared bar earlier and it wasn't working for me... 02:21
It's sad that something like type coercions isn't in perl6...I loved that part of moose 02:22
sorear it's sad that perl 6 hasn't bothered to learn from the changes Moose made after being forked 02:24
perl 6 doesn't even have _required attributes_
Lard yeah I saw the = die "blah" but I prefer the declarative moose style 02:25
sorear eh. I shouldn't dump all my complaints on the new person.
Lard haha 02:26
sorear weclome! 02:29
Lard thanks :)
pmichaud gist.github.com/3063969 # does this work as a Foo($value) coercer? 02:33
then we could have my Foo() $x; such that any value assigned to $x passes through Foo(...) first 02:34
although I guess the resulting value would need to be blessed into Foo to satisfy the type constraint. 02:35
sorear let's add to Mu/Any multi postcircumfix:<( )>($x) { $x."{self.^name}"() } 02:36
now you can implement either Type() or .Type, with Type() multis taking precedence
although we don't currently have a good way to define a ."Foo::Bar" method in Perl 6 02:37
pmichaud I'm guessing you really want multi postcircumfix:<( )>(Mu:U: $x) there, though.
otherwise we're defining postcircumfix on... well, everything :-) 02:38
sorear hrm. 02:43
irclog.perlgeek.de/perl6/2010-07-16#i_2559776
I recollected this as being much more of unilateral shoot-down than it actually was 02:44
quietfanatic Hm, is there a way to do a (function-style) pattern match on an object with attributes? 03:40
I know you can do it with an array.
perl6: sub sayattrs ([*%attrs]) { say %attrs.perl }; class Vec { has $.x; has $.y; }; sayattrs(Vec.new(x => 1, y => 2)) 03:41
p6eval niecza v19-12-gf36d743: OUTPUT«Unhandled exception: Nominal type check failed in binding '' in 'sayattrs'; got Vec, needed Positional␤ at /tmp/L__dOahayc line 0 (sayattrs @ 1) ␤ at /tmp/L__dOahayc line 1 (mainline @ 7) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3918 (ANON @ 3) ␤ …
..rakudo 57eaaa: OUTPUT«Nominal type check failed for parameter ''; expected Positional but got Vec instead␤ in sub sayattrs at /tmp/MZ8UbWdvP6:1␤ in block <anon> at /tmp/MZ8UbWdvP6:1␤␤»
quietfanatic perl6: sub sayattrs ({*%attrs}) { say %attrs.perl }; class Vec { has $.x; has $.y; }; sayattrs(Vec.new(x => 1, y => 2))
p6eval niecza v19-12-gf36d743: OUTPUT«===SORRY!===␤␤Unable to parse signature at /tmp/HiNGWvDQHD line 1:␤------> sub sayattrs (⏏{*%attrs}) { say %attrs.perl }; class Ve␤Couldn't find final ')'; gave up at /tmp/HiNGWvDQHD line 1:␤------> sub sayattrs (…
..rakudo 57eaaa: OUTPUT«===SORRY!===␤Missing block␤at /tmp/i4BtaLswMt:1␤»
sorear r: sub sayattrs($ (*%attrs)) { say %attrs.perl }; class Vec { has $.x; has $.y; }; sayattrs(Vec.new(x => 1, y => 2)) 03:43
p6eval rakudo 57eaaa: OUTPUT«("x" => 1, "y" => 2).hash␤»
sorear quietfanatic.
quietfanatic oh 03:44
sweet
though the syntax is a little odd methinks
Oh, but I can't use that to do coercions, because methods on SomeType:U appear to be impossible 03:50
rats
Oh never mind, I was spelling postcircumfix 'infix'. Doh 03:52
Except it still doesn't work. 03:53
class A { method postcircumfix:<( )> (A:U:) { say 3 } }; A()
rn: class A { method postcircumfix:<( )> (A:U:) { say 3 } }; A()
p6eval rakudo 57eaaa: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤ in method postcircumfix:<( )> at /tmp/bxmz9cNgsh:1␤ in <anon> at src/gen/BOOTSTRAP.pm:811␤ in any <anon> at src/gen/BOOTSTRAP.pm:807␤ in block <anon> at /tmp/bxmz9cNgsh:1␤␤»
..niecza v19-12-gf36d743: OUTPUT«3␤»
quietfanatic niecza++ 03:54
pmichaud r: class XYZ { method postcircumfix:<( )>(XYZ:U: $x) { say 3 } }; XYZ(0) # checking
p6eval rakudo 57eaaa: OUTPUT«3␤»
pmichaud r: class XYZ { method postcircumfix:<( )>(XYZ:U: $x) { say 3 } }; XYZ() # checking
p6eval rakudo 57eaaa: OUTPUT«3␤»
pmichaud hmmmm
r: class XYZ { method postcircumfix:<( )>(XYZ:U: $x) { say 3 } }; XYZ(3,4,5) # checking 03:55
p6eval rakudo 57eaaa: OUTPUT«3␤»
pmichaud I bet Rakudo is always expecting a Capture of the arguments there.
quietfanatic huh
pmichaud r: class XYZ { method postcircumfix:<( )>(XYZ:U: $x) { say $x.WHAT } }; XYZ(3,4,5) # checking
p6eval rakudo 57eaaa: OUTPUT«Capture()␤»
pmichaud *ding*
quietfanatic Where does that capture come from? 03:56
pmichaud it's the capture formed by the arglist to the .( ) call
quietfanatic oh, so 03:57
r: class XYZ { method postcircumfix:<( )>(XYZ:U: $x) { say $x[1] } }; XYZ(3,4,5) # checking
pmichaud I'm not sure it's _correct_; that just appears to be what Rakudo is doing.
p6eval rakudo 57eaaa: OUTPUT«4␤»
quietfanatic er
r: class XYZ { method postcircumfix:<( )>(XYZ:U: $x) { say $x } }; XYZ(3,4,5) 03:58
p6eval rakudo 57eaaa: OUTPUT«3 4 5␤»
quietfanatic yeah I'm pretty sure that's wrong.
where's masakbot?
sorear quietfanatic: 3,4,5 are being wrapped into a capture. 03:59
quietfanatic Yes, but the signature should unwrap the capture.
pmichaud there may be some circularity sawing taking place there... probably need to check with jnthn++, too. 04:00
dalek p: b9d44ce | pmichaud++ | src/QRegex/ (3 files):
[qregex] Initial implementation of <( and )> tokens.
 04:20
p: 66edbf6 | pmichaud++ | t/qregex/01-qregex.t:
Add some got/expected output to t/qregex/01-qregex.t .
p: b07bc9e | pmichaud++ | t/qregex/rx_metachars:
[qregex]: Add some tests for <( and )> .
moritz \o 04:31
pmichaud o/
sorear |o|
pmichaud eek! a TIE fighter!
dalek kudo/nom: a569e39 | pmichaud++ | / (4 files):
Update <( and )> in regexes to use code inherited from NQP.
 04:34
ast: baf17a4 | pmichaud++ | S05-metasyntax/angle-brackets.t:
Add some more <( and )> tests.
tadzik hoelzro|away: yeah, please don't worry about licesing when it comes to my modules :) 06:48
I missed some action due to being suddenly asleep 06:49
sorear o/ tadzik 06:53
tadzik sorear: o/ 06:56
moritz \o 07:02
sergot hello o/ 07:47
!
sorear helo sergot 07:59
szabgab r: say "hi" 08:28
p6eval rakudo a569e3: OUTPUT«hi␤»
szabgab r: for 1 .. 3 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤2␤3␤»
szabgab r: for 1 .. 3 { say }
p6eval rakudo a569e3: OUTPUT«␤␤␤»
szabgab r: for 1 .. 3 { print }
p6eval rakudo a569e3: ( no output )
szabgab the last one prints me 3 times True in the REPL while the say version just 3 empty rows 08:29
If anyone could explain that to me ... 08:30
pmichaud the repl only does the 'p' if the executed code doesn't produce any output
so, since for 1..3 { say } generates output, the repl doesn't do any additional printing.
for 1..3 { print } doesn't print anything (it prints 3 null strings), so the repl prints the result of the for loop 08:31
which is (True, True, True)
r: say (for 1..3 { print }).perl
p6eval rakudo a569e3: OUTPUT«(Bool::True, Bool::True, Bool::True).list␤»
pmichaud you can also see the difference if you do say "" and print "" from the repl 08:32
the first will display a blank line; the second will display "True" (which is the result of the print)
szabgab ok for that, and why neiher of them print the content of $_ ?
pmichaud there are no more default arguments in Perl 6.
pmichaud if you want to print $_, you have to do print $_ or $_.print or .print 08:33
szabgab but there is for the for loop
default iterator
or what is the name of the variable there?
pmichaud yes, $_ is set in the block, but a bare "say" doesn't default to displaying $_
std: for 1..3 { say } # I think STD flags it as an error 08:34
p6eval std fd2647b: OUTPUT«===SORRY!===␤Unsupported use of bare 'say'; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argument at /tmp/BRcBVzOyPz line 1:␤------> for 1..3 { say⏏ } # I think STD flags it as an error␤Check failed…
pmichaud we used to flag it as an error in Rakudo... apparently that got dropped somewhere.
szabgab nice
szabgab probably it would be better to flag that too 08:34
pmichaud agreed; it's not always super-easy to do that :)
szabgab and what is the name of the variables in the for loop ?
pmichaud the for loop just passes the iterated values to the block -- the block gets to decide what to call the variable. 08:35
r: for 1..3 { say $^x }
p6eval rakudo a569e3: OUTPUT«1␤2␤3␤»
pmichaud r: for 1..3 -> $n { say $n }
p6eval rakudo a569e3: OUTPUT«1␤2␤3␤»
szabgab I mean I used to call them "iterator"
in perl 5
to describe them
in general
pmichaud I don't think that works in p6, though.
because they belong to the block, not the for loop. 08:36
r: (1..3).map( -> $n { say $n }).eager
p6eval rakudo a569e3: OUTPUT«1␤2␤3␤»
szabgab so I was looking for a generic term for the variable that is on the right hand side of ->
pmichaud it's a parameter 08:37
szabgab ok, that will do :)
pmichaud just like if we do sub anon ($x, $y) { ... }
szabgab then another question , there used to be an idea to write for 1..10 by(2) { }
pmichaud we can also do -> $x, $y { ... }
szabgab but I don't think it is there any more, not even in the spec 08:38
pmichaud correct, it's gone. We now use the sequences operator. 08:38
szabgab so can I loop every 2nd value?
pmichaud say 1,3,5 ...^ *>= 10
szabgab r: for 1,3...10 { say $_ }
pmichaud r: say 1,3,5 ...^ *>= 10~
p6eval rakudo a569e3: OUTPUT«(timeout)»
rakudo a569e3: OUTPUT«===SORRY!===␤Confused␤at /tmp/8DXq5AQqdk:1␤»
pmichaud r: say 1,3,5 ...^ *>= 10
p6eval rakudo a569e3: OUTPUT«1 3 5 7 9␤»
szabgab r: for 1,3,5 ...10 { say $_ }
p6eval rakudo a569e3: OUTPUT«(timeout)» 08:39
pmichaud the endpoint has to match exactly, or the sequence is infinite
szabgab r: for 1,3,5 ...11 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤11␤»
szabgab r: for 1,3 ...11 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤11␤»
szabgab oh
pmichaud it's a smart match at the end of the sequence, thus * >= 10 works 08:39
szabgab I think that used to work without exact match
pmichaud it did; that got changed
because we have lots of sequences that can't be easily deduced as "up to" or "down to" 08:40
szabgab r: for 1,3 ...^ *>10
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Missing block␤at /tmp/N0PJ14uCw4:1␤»
szabgab r: for 1,3 ...^ *>10 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤»
pmichaud hmmm, I wonder 08:42
szabgab r: for 1,2,0 ... 3 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤2␤0␤Unable to deduce sequence␤ in method gist at src/gen/CORE.setting:9567␤ in sub say at src/gen/CORE.setting:7011␤ in block <anon> at /tmp/Dx2zwGalQA:1␤␤»
pmichaud r: for 1,3 ...^ * !< 10 { .say }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤»
pmichaud heh, it works.
szabgab could you give a sequence where you cannot deduct if it is upper or lower limit? 08:43
pmichaud r: my @a = 1, -2, 4 ... *; say @a[^15]; 08:44
p6eval rakudo a569e3: OUTPUT«1 -2 4 -8 16 -32 64 -128 256 -512 1024 -2048 4096 -8192 16384␤»
pmichaud r: my @a = 1, -1, 1 ... *; say @a[^10]
p6eval rakudo a569e3: OUTPUT«1 -1 1 -1 1 -1 1 -1 1 -1␤»
szabgab r: for 1,2,0 ... 3 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤2␤0␤Unable to deduce sequence␤ in method gist at src/gen/CORE.setting:9567␤ in sub say at src/gen/CORE.setting:7011␤ in block <anon> at /tmp/1Mj25vrSPb:1␤␤»
pmichaud r: my @a = True, !* ... *; say @a[^10] 08:45
p6eval rakudo a569e3: OUTPUT«True False True False True False True False True False␤»
szabgab it does not lie my sequence :(
like
pmichaud it can't figure out the relationship between 1, 2, and 0
szabgab r: for 1,2,0,3 ... -1 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤2␤0␤3␤Unable to deduce sequence␤ in method gist at src/gen/CORE.setting:9567␤ in sub say at src/gen/CORE.setting:7011␤ in block <anon> at /tmp/8yQEiguzOR:1␤␤»
pmichaud what sequence are you wanting there? 08:46
szabgab +1, -2, +3, -4
pmichaud yeah, that's neither arithmetic nor geometric
but you can do
(thinking) 08:47
diakopter 1234 times -1,1,-1,1
the nth at the first times the nth at the second 08:48
szabgab I was thinking that if it is restricted to arithmetic and geometric then from 3 items you can see if it has a direction
pmichaud r: my @a = 1, { $^x.abs * -$^x.sign } ... *; say @a[^10]
p6eval rakudo a569e3: OUTPUT«1 -1 1 -1 1 -1 1 -1 1 -1␤»
pmichaud r: my @a = 1, { ($^x.abs+1) * -$^x.sign } ... *; say @a[^10]
p6eval rakudo a569e3: OUTPUT«1 -2 3 -4 5 -6 7 -8 9 -10␤»
szabgab and from there you can no if it is upper bound, lower bound or no bound
pmichaud szabgab: yeah, I think we decided that's too much special casing
it's much easier to just say it's always a smartmatch at the end 08:49
and geometric doesn't always have a direction 08:50
szabgab r: for 1,3 ...^ *<10 { say $_ }
p6eval rakudo a569e3: ( no output )
szabgab r: for 1,3 ...7 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤»
szabgab r: for 1,3 ...^ *>10 { say $_ } 08:51
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤»
szabgab it feels the oposite way :)
pmichaud yeah, that's why I did !< earlier :-)
szabgab r: for 1,3 ...^ !<10 { say $_ }
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix !< instead␤at /tmp/GihLfh61fo:1␤»
pmichaud * !< 10
szabgab r: for 1,3 ...^ * !<10 { say $_ } 08:52
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤»
szabgab cryptic :(
pmichaud not really.... !< is "not less than"
szabgab r: for 1,3 ... ^*!< 10 { say $_ }
p6eval rakudo a569e3: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \$v, Mu *%_)␤␤ in method Numeric at src/gen/CORE.setting:686␤ in sub prefix:<^> at src/gen/CORE.setting:4948␤ in block <anon> at /tmp/bagpRPf6fq:1␤␤»
szabgab r: for 1,3 ...^*!< 10 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤»
szabgab 4 non-alpha :) 08:53
r: for 1,3 ...* < 10 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤»
szabgab r: for 1,3 ...* > 10 { say $_ }
p6eval rakudo a569e3: OUTPUT«1␤3␤5␤7␤9␤11␤»
szabgab the first one would look good :) 08:54
oh this will never end if every person comes with new ideas :)
or "different from current" even if not "new" 08:55
pmichaud r: say 'abcd'.match(/a/, :x(1..4)).WHAT 09:39
p6eval rakudo a569e3: OUTPUT«List()␤»
pmichaud r: say 'abcd'.match(/a/, :x(1..4))
p6eval rakudo a569e3: OUTPUT«a␤»
pmichaud r: say 'abcd'.match(/a/, :n(1,3))
p6eval rakudo a569e3: OUTPUT«q[a]␤␤»
pmichaud r: say 'abcd'.match(/a/, :n(1,3)).WHAT
p6eval rakudo a569e3: OUTPUT«Match()␤»
pmichaud r: say 'abcd'.match(/./, :n(1,3)).WHAT 09:40
p6eval rakudo a569e3: OUTPUT«Match()␤»
pmichaud r: say 'abcd'.match(/./, :nth(1,3))
p6eval rakudo a569e3: OUTPUT«a c␤»
pmichaud r: say 'abcd'.match(/./, :nth(1,3)).WHAT
p6eval rakudo a569e3: OUTPUT«List()␤»
pmichaud r: say 'abcd'.match(/a/, :nth(1,3)).WHAT
p6eval rakudo a569e3: OUTPUT«Match()␤»
pmichaud that seems wrongish. 09:40
time for sleep -- bbl 09:43
Su-Shee hi all :) 09:51
sergot Su-Shee: hi 09:54
masak g'day, #perl6. 10:25
colomon \o 10:25
colomon is not quite awake yet
masak is on a bus!
and I've been awake since 04:30, which... mhh.
colomon work up at exactly 06:01 10:27
tried to go back to sleep, but my nose was too stuffy to handle the CPAP. 10:29
diakopter hm. I spupose I should sleep tonight
jnthn Good Saturday, #perl6 10:42
masak jnthn! \o/
moritz \o 10:44
jnthn++ # hague grant completed
colomon \o/ 10:46
jnthn \o/
masak jnthn++ \o/ 10:56
masak re irclog.perlgeek.de/perl6/2012-07-07#i_5793473 -- I agree with quietfanatic++, and I think sorear++ changed the spec and fixed Niecza to be more consistent and simpler in this regard. I don't believe there are circularity issues. 11:06
if jnthn concurs (or doesn't fly into a tantrum about having to change this bit), I'll submit a rakudobug.
jnthn It used to be spec'd that way. 11:07
I agreed it could change already.
Rakudo didn't get updated yet.
It was mostly SMOP that wanted it to be the other way, fwiw.
masak submits rakudobug 11:08
jnthn That particular circularity saw doesn't apply to anything 6model-y. 6model doesn't care for captures in any sense, and the invocation protocol is specified as VM specific.
masak yes, I remember that bit of input coming from pmurias++ and ruoso++
jnthn Rakudo's postcircumfix:<( )> isn't the invocation protocol in general.
SMOP tried to make it that way, which was why it wanted said circularity saw, iirc.
.oO( Can I just fix this, or do I have to think about deprecation policy... :-) )
 11:09
masak the invocation protocol is VM specific, but I do hope the expected parameter list for postcircumfix:<( )> isn't. 11:14
jnthn masak: That was kinda my point. What postcircumfix:<( )> does isn't anything to do with the VM's view. 11:15
masak ok ok good.
that's probably wise.
masak Woodi! \o/ 11:39
pmichaud jnthn: (postcircumfix:<( )>) afaic, you can "just fix it". 11:51
Woodi hi today everyone :) 11:52
jnthn as far as i...? 11:53
pmichaud as far as I'm concerned
jnthn is used to that ending "ct" = can tell :)
oh!
:)
dalek c: 2139f4f | pmichaud++ | lib/Positional.pod:
[Positional]: EnumMap isn't Positional; other minor tweaks and fixes.
 12:18
dalek c: 7b0caa5 | pmichaud++ | lib/ (7 files):
Various grammatical and minor fixes.
 12:31
kudo/toqast: f96df3b | jnthn++ | src/QPerl6/Actions.pm:
Translate the rest of the 'obvious' remaining PAST:: usages in the Actions. Thsoe that remain are interesting in some way. :-)
 12:33
kudo/toqast: 0c943c2 | jnthn++ | src/QPerl6/World.pm:
Translate the easy remaining PAST usages in World.
dalek p/toqast: fd47dae | jnthn++ | src/QAST/Operations.nqp:
Add what/how/who ops.
 13:05
p/toqast: 58acfd1 | jnthn++ | src/QAST/ (2 files):
Add a way to mark a block as having a custom arguments processor, and compile it to a :call_sig usage on Parrot.
kudo/toqast: 9b02aca | jnthn++ | src/QPerl6/ (3 files):
Add a few ops to the abstraction layer.
 13:07
kudo/toqast: c62d139 | jnthn++ | src/QPerl6/Actions.pm:
Update arguments handling. Gets back one more sanity test, and changes the failure mode of many others.
kudo/toqast: bf653dd | jnthn++ | src/QPerl6/Actions.pm:
Update compilation of .WHAT/.HOW/.WHO.
dalek kudo/toqast: 9114f92 | jnthn++ | src/QPerl6/ (2 files):
Start to update return value handling.
 13:22
kudo/toqast: a543d24 | jnthn++ | src/QPerl6/ (2 files):
Update Perl 6 capture_lex handling, which wins back one more sanity test.
kudo/toqast: 3f5bb89 | jnthn++ | src/QPerl6/Actions.pm:
Make returnless analysis not explode.
timotimo are there already specs or ideas how the automatic commandline argument parser created with MAIN can be enhanced? descriptions for flags and values etc.? 13:58
moritz not really 14:12
you can always write your own command line parser
timotimo it would be nice to have a nice default one in place that does more than just the simples stuff
moritz then write one 14:13
and we'll ship it with R*
and with a bit of luck and work, it'll become the default
colomon timotimo: do you know what the current one can do?
timotimo i don't know all of it i'm sure
moritz it's nice, but it does have its limitations 14:14
dalek p/toqast: 9c4dfa5 | jnthn++ | / (4 files):
Implement and test resultchild, which allows something other than the final statement of a QAST::Stmt or QAST::Stmts to be used as the result of the overall operation.
moritz it doesn't support subcommands with options separate from the main options, for example
colomon What's qx// supposed to do if it can't execute? (Say you've used the wrong command name.) 14:21
moritz fail() I hope
colomon that seems reasonable
flussence that's what S29 says for shell() and run() anyway 14:22
moritz the first two tests in t/00-parrot/ now pass with qperl6 14:25
as does no. 4
jnthn moritz: Currenlty there's 4 failures in t/00-parrot nad 5 in t/01-sanity 14:26
(With latest toqast in nqp and rakudo repos)
At least, that's how I have it :)
colomon arg. in niecza, sprintf works (on my current case) but printf doesn't. :( 14:33
niecza: printf("%s, %d\n", "test", 1) 14:35
p6eval niecza v19-12-gf36d743: OUTPUT«Unhandled exception: index out of range␤ at /home/p6eval/niecza/lib/CORE.setting line 1277 (sprintf @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1278 (printf @ 4) ␤ at /tmp/R2j1ncpN6z line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setti…
_sri wonders if there's a perl6brew yet
tadzik: i've been looking at MuEvent, is $socket.poll the only way to do non-blocking I/O? 14:41
that approach doesn't scale at all 14:42
colomon is not sure what it says about him that fixing bugs in Niecza is easier than just remembering the syntax to get his 16-line $work script working in p5 14:43
_sri wanted to build a little event loop with rakudo to see where it stands performance wise for I/O heavy servers 14:46
dalek p/toqast: 2cf193e | jnthn++ | src/QAST/Operations.nqp:
Port lexotic to QAST.
 14:55
dalek volaj: 8449183 | (Arne Skjærholt)++ | TODO:
Update TODO list.
 15:12
dalek ecza: 442e075 | (Solomon Foster)++ | lib/CORE.setting:
Make printf work. (How had we missed this one?!)
 15:20
dalek p/toqast: f80052a | jnthn++ | src/QAST/Compiler.nqp:
s/newclosure/capture_lex/ for consistency with PAST.
 16:00
p/toqast: 7de611e | jnthn++ | src/QAST/ (2 files):
Fix register allocation so that lexicals don't get stomped on.
kudo/toqast: 05a1e61 | jnthn++ | src/QPerl6/Actions.pm:
Update return handling for QAST.
kudo/toqast: c6c6488 | jnthn++ | src/QPerl6/Actions.pm:
Stop add_inlining_info blowing up, though really this whole thing goes away soon since we'll be able to serialize bits of QAST.
p/toqast: 2cb23aa | jnthn++ | src/QAST/Compiler.nqp:
Ensure blocks always end up with a lexinfo.
 16:16
harmil_home busy morning 16:19
ugh, I'm trying to refactor IO into a base IO::Parrot and the specced IO class, and I'm getting some very odd errros on compile that give me no leads on what to fix... 16:21
moritz harmil_home: which specced IO class? 16:22
harmil_home: and what errors do you get?
harmil_home moritz: I don't know. Whatever one was already in there. I'm basically trying to leave it alone, and just factor out $!PIO into a base class
Could not locate compile-time value for symbol Comp::AdHoc 16:23
current instr.: 'nqp;Perl6;World;_block2280' pc 32075 (src/gen/perl6-symboltable.pir:13915) (src/Perl6/World.pm:1709)
moritz harmil_home: that IO class isn't specced :-)
harmil_home moritz: fair enough, but it's loosely based on (and probably intended to migrate towards) the spec, right?
moritz harmil_home: very loosely 16:24
harmil_home My goal is to leave its interface as its interface, whatever that is and make Parrot::IO give us everything Parrot can do
er IO::Parrot
moritz ah, I see how you might get that error 16:25
harmil_home So, for example, my buffered IO stuff just moves down a level, as does anything that's basically just a wrapper around nqp::p6box_*($!PIO.somefunc())
oh?
moritz harmil_home: it will be much easier for you do the whole thing outside of the setting 16:26
in fact, there isn't a good reason to do it inside the setting 16:27
unless you are fond of very long compile times
harmil_home I'm not entirely clear what that menas.
moritz just write the new IO::Parrot and IO class in a normal script
not in src/core/IO.pm
harmil_home Ah, and override the core ones and all of the static functions that map into them? 16:28
moritz (the error you are getting is an error during error reporting; I'll try to make it more robust)
harmil_home: right
whever "override" just means "declare a new one"
(lexical lookup)++ 16:29
harmil_home I suppose I could do that, but it seems like a lot of duplication. Won't that lead to IO in the core slowly diverging from "the real IO"?
Do we care?
moritz harmil_home: well, we can always copy it back into src/core/IO.pm if it proves to be superior
harmil_home Fair enough, I see what you're saying. 16:30
moritz harmil_home: in fact I prototyped most of "my" setting code outside the setting first
arnsholt One of the benefits of having most of the compiler written in Perl 6: Prototyping new bits of the compiler can often be done outside of the compiler first =) 16:32
pmichaud Yes, I almost always prototype new features outside of the core setting first. 16:33
question: what should be the result of
harmil_home Should I do this as a stand-alone perl6-io module or in my fork of rakudo/rakudo under lib?
pmichaud "abcd".match( /./, :p(1), :global) 16:34
harmil_home: I suggest a standalone module.
harmil_home pmichaud: thanks
pmichaud I think (1) IO::Parrot is very useful, yet (2) it's unlikely to be part of the Rakudo core, unless we find it to be fundamental somehow. 16:35
(it is likely to be part of R*, however)
rn: say "abcd".match( /./, :p(1), :global) 16:36
harmil_home pmichaud: urm... well, I can't do IO::Parrot in a vacuum. IO::Parrot is mostly a re-factor and re-write of the internals of IO
p6eval niecza v19-13-g442e075: OUTPUT«Unhandled exception: Excess arguments to Cool.match, unused named global␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (Cool.match @ 1) ␤ at /tmp/AYDLrnj5tV line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3918 (ANON @ 3) ␤ at /ho…
..rakudo a569e3: OUTPUT«b c d␤»
pmichaud rn: say "abcd".match( /./, :p(1), :g)
p6eval niecza v19-13-g442e075: OUTPUT«Unhandled exception: Excess arguments to Cool.match, unused named g␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (Cool.match @ 1) ␤ at /tmp/ePKgPPnjAJ line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3918 (ANON @ 3) ␤ at /home/p6…
..rakudo a569e3: OUTPUT«b c d␤»
moritz well, :p only makese sense on the first match 16:36
pmichaud yes, so I'm wondering if we should only get one match back. 16:37
moritz I can see three possible interpretations 16:37
(1) only ever return one match
(2) only anchor the first match
moritz (3) anchor the first match to 1, and each subsequent match to where the previous left off 16:38
pmichaud yes, those are the three I was considering also. I guess this will have to be a TimToady++ question.
arnsholt My initial reaction is 1, I think
moritz nqp: my $x := 1; $x++; say($x) 16:39
p6eval nqp: OUTPUT«2␤»
moritz nqp: my $x := 1; ++$x; say($x)
p6eval nqp: OUTPUT«2␤»
pmichaud github.com/perl6/specs/issues/16 16:57
moritz pmichaud++ 16:58
sorear: can we get dalek to report spec issues? 16:59
pmichaud I found that by just watching the perl6/specs repo, I get emails everytime a new issue or comment is posted.
email seems to be the only push notification for issues on github 17:00
moritz well, dalek also supports pulling
colomon ah, that's why I've been getting those emails...
moritz having commit access should also be enough
pmichaud I think one has to actually be watching the repo to get the emails. 17:01
but I could be wrong about that.
(You also get emails for any issues you create or comment upon, I think.)
moritz maybe I'm watching it because I created it :-) 17:02
I don't remember ever pressing a "watch" button though
pmichaud anyway, I looked into it a few days ago, and decided I wasn't up to tracking everything down to get issue updates into dalek 17:08
dalek p/toqast: ad562d8 | jnthn++ | src/QAST/Operations.nqp:
Implement op xor (which has the semantics of xor_nqp).
 17:12
ecs: aca4dff | pmichaud++ | questions.txt:
Remove questions.txt in favor of Github's issue tracker.

  github.com/perl6/specs/issues/17 .
kudo/toqast: 2ce8e79 | jnthn++ | src/QPerl6/Grammar.pm:
xor_nqp => xor; now down to 2 failing test files in 00-parrot and 3 in 01-sanity.
 17:13
harmil_home rn: class A {}; role B {}; class C is A does B {} 17:17
szabgab r: say substr "abcd", 1, -1
p6eval rakudo a569e3, niecza v19-13-g442e075: ( no output )
rakudo a569e3: OUTPUT«Negative length argument (-1) to .substr␤ in method gist at src/gen/CORE.setting:9567␤ in sub say at src/gen/CORE.setting:7011␤ in block <anon> at /tmp/kvgD4R6vnP:1␤␤»
harmil_home I think the no output is mine ;) 17:18
szabgab is the fact that substr does not take negative length the design or lack of implementation?
pmichaud I think it's *-1
harmil_home rn: class A {}; role B {}; class IO is A does B
p6eval niecza v19-13-g442e075: OUTPUT«===SORRY!===␤␤Unable to parse class definition at /tmp/OcrdwU6Zc9 line 1 (EOF):␤------> ss A {}; role B {}; class IO is A does B⏏<EOL>␤␤Parse failed␤␤»
..rakudo a569e3: OUTPUT«===SORRY!===␤Unable to parse $*PKGDECL definition at line 2, near ""␤»
szabgab harmil_home: yes
pmichaud r: say substr "abcd", 1, *-1
p6eval rakudo a569e3: OUTPUT«bc␤»
szabgab pmichaud: aha 17:19
thanks
pmichaud the error message to .substr should perhaps indicate that.
harmil_home anyone have a sense of why I'm getting errors parsing a class definition if I call it IO?
szabgab r: my $x=-1; say substr "abcd", 1, $x
p6eval rakudo a569e3: OUTPUT«Negative length argument (-1) to .substr␤ in method gist at src/gen/CORE.setting:9567␤ in sub say at src/gen/CORE.setting:7011␤ in block <anon> at /tmp/DgW1qkrzzW:1␤␤»
moritz r: class IO { }
p6eval rakudo a569e3: ( no output )
szabgab r: my $x=-1; say substr "abcd", 1, *$x 17:20
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Confused␤at /tmp/oj8gSkFa0S:1␤»
moritz harmil_home: no. works here
szabgab r: my $x=-1; say substr "abcd", 1, * $x
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Confused␤at /tmp/Y2_1LUWr3z:1␤»
szabgab r: my $x=1; say substr "abcd", 1, * -$x
p6eval rakudo a569e3: OUTPUT«bc␤»
harmil_home moritiz, see my rn: above
moritz, even
szabgab so both rakudo and I am confused :)
pmichaud rn: class A {}; role B {}; my class IO is A does B { ... } 17:23
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤The following packages were stubbed but not defined:␤ IO␤at /tmp/7bbkQWsjSF:1␤»
..niecza v19-13-g442e075: OUTPUT«===SORRY!===␤␤Package was stubbed but not defined at /tmp/4sq6dqLtRv line 1:␤------> class A {}; role B {}; my class⏏ IO is A does B { ... }␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.se…
pmichaud rn: class A {}; role B {}; my class IO is A does B { }
p6eval rakudo a569e3, niecza v19-13-g442e075: ( no output )
pmichaud rn: class A {}; role B {}; class IO is A does B { }
p6eval rakudo a569e3, niecza v19-13-g442e075: ( no output )
pmichaud seems to work.
harmil_home I'm trying to figure out what you did differently... 17:24
moritz harmil_home: you were missing the { } of the class 17:25
harmil_home ah... hmm...
so then it was a red herring. That's unpleasant
My real error is Unable to parse $*PKGDECL definition at line 226, near "does IO::F" 17:26
which I get on class IO is IO::Parrot does IO::FileTestable {
pmichaud could be an error inside the braces (that isn't being reported well)
moritz r: class IO::Parrot { }; class IO is IO::Parrot does IO::FileTestable { } 17:27
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Unable to parse $*PKGDECL definition at line 2, near "does IO::F"␤»
harmil_home OK, I'll start doing the binary search...
pmichaud if you can nopaste it, we might be able to spot it quickly
harmil_home interesting.
moritz seems I've just reproduced it above
r: class IO::Parrot { }; class MyIO is IO::Parrot does IO::FileTestable { }
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Unable to parse $*PKGDECL definition at line 2, near "does IO::F"␤»
moritz so, not related to the IO name
harmil_home related to undefined IO::FileTestable? 17:28
moritz r: class IO::Parrot { }; class IO does IO::FileTestable { }
p6eval rakudo a569e3: OUTPUT«===SORRY!===␤Unable to parse $*PKGDECL definition at line 2, near "does IO::F"␤»
harmil_home Oh heck
I didn't export IO::Parrot after I moved it out into a file
moritz r: say IO::FileTestable
p6eval rakudo a569e3: OUTPUT«Could not find symbol 'IO::&FileTestable'␤ in block <anon> at /tmp/ojDGWQZdeJ:1␤␤»
moritz oh, that explains a lot
harmil_home hmmm that didn't help any 17:29
pmichaud might be worthy of a LTA rakudobug
moritz r: say IO::FileTestable
p6eval rakudo a569e3: OUTPUT«Could not find symbol 'IO::&FileTestable'␤ in block <anon> at /tmp/k8rQXz5Tio:1␤␤»
harmil_home LTA?
moritz less than awesome
harmil_home ah
moritz oh
that's becase p6eval hides IO
it works locally
but yes, a missing IO::Parrot would do the same 17:30
pmichaud also, if you do class IO does IO::FileTestable you're hiding the old IO namespace
harmil_home pmichaud, which I wish to do
I'm wholly replacing it
pmichaud which means that IO::FileTestable isn't visible :) 17:31
unless you're replacing that too :)
harmil_home yep
pmichaud okay then.
moritz though, consider not to inherit from IO::Parrot
harmil_home my IO.pm is the core IO.pm with all parrot IO calls ripped out
moritz it feels wrong to have VM-specific behavior as the superclass of the general IO class 17:32
harmil_home how would you define IO.read if not in terms of a lower-level interface?
pmichaud it can be a role 17:33
jnthn It's fine to define it in terms of a lower level interface, but it should be done through delegation.
moritz harmil_home: you can ouse a lower-level interface without inheritance
jnthn Please think a bit about what this will look like off Parrot too. :) 17:34
moritz harmil_home: it can be an attribute (has $.parrot-io), or simply function calls or something else
right, delegation, not inheritance
pmichaud we definitely don't want to be interposing a vm-specific class between IO and Any
harmil_home Ugh, I was trying to extract the parrot-speciifc attribute out, not change it's type ;)
pmichaud, why not?
pmichaud it is likely to confuse the type system across implementations 17:35
moritz because superclasses are part of the public API
pmichaud moritz++ said it better :)
moritz and our public API shouldn't expose VM-specific stuff
inheritance is just wrong Wrong WRONG here
harmil_home pmichaud, how would it confuse the type systems across implementations? There won't be an IO::Parrot in a non-Parrot implementation will there?
pmichaud harmil_home: correct. Which means that IO on a different implementation would have a different superclass than Rakudo on Parrot 17:36
harmil_home sigh, OK, I'm going to go get some lunch and try to pretend that getting buffered IO into the core isn't running into road-blocks at every step
moritz be glad that we caught that mistake *before* you implemented it all 17:38
pmichaud > say substr 'abcd', 1, -1 17:40
Negative length argument (-1) to .substr, use '*-1' if you want offset from the end
jnthn pmichaud++
dalek kudo/toqast: 3dd5f6f | jnthn++ | src/QPerl6/Actions.pm:
Eliminate a usage of viviself.
kudo/toqast: a466813 | jnthn++ | src/QPerl6/ (2 files):
Start fixing package lookups a bit. Gets us one more test passing again.
jnthn splice still needs this kinda update too
sorear good * #perl6 17:46
moritz: dalek reports anything that you can append to the queue file. 17:47
moritz: it also has an rss puller
rss pull will be slower ofc
moritz sorear: where is parrot's issue watching configured? 17:48
pmichaud ...parrot has an issue watcher? 17:49
pmurias sorear: hi
sorear: is there anything with niecza that you need help with? 17:50
sorear pmichaud: yes! and it even used to work! 17:53
trac only
sorear colomon: why doesn't sprintf take a slurpy? 17:56
dalek kudo/nom: c24da0b | pmichaud++ | src/core/Str.pm:
Improve the negative start/length failure messages for substr.
 18:01
pmurias sorear: my summer break has started so I should have lots of free time soon 18:05
sorear: should I work on the stuff from the TODO or does the no-bootstrap branch need to be finished first? 18:06
sorear I haven't gone over TODO recently, probably it has a lot of stale items 18:09
I personally am blocking everything else on no-bootstrap
pmurias anything there I should work on 18:13
? 18:14
preferably something isolated as lightning destroyed my dsl modem, so I have to walk to the uni to use the wifi :(
sorear shrugs 18:21
pmurias sorear: how should I build the non-bs branch? 18:54
sorear pmurias: make, then mono obj/Run.Kernel.dll -test-nobs --help 18:58
hmmm
no, that was right 18:59
it wasn't working because I tried it in the wrong checkout
flussence random question: is there a more idiomatic way to have a repeating list of items than @f[$_ % @f]? 19:00
sorear @f xx *
pmurias sorear: thanks 19:01
sorear: I mistook the lack of some debian package for typing the wrong invocation
flussence oh, that looks nicer. thanks
sorear pmurias: actually just make obj/Run.Kernel.dll 19:02
the branch is nowhere near usable
pmurias sorear: does the non-bs branch run anything right now? 19:07
sorry for repetition 19:08
sorear no
pmurias so you are just translating code for now? 19:10
sorear yes 19:11
masak evening, #perl6 19:16
sorear o/ masak
moritz \o * 19:25
masak r: class A { method foo() {} }; say +grep { .positional && !.invocant }, A.^methods[0].signature.params; class B { has A $.a handles <foo> }; say +grep { .positional && !.invocant }, B.^methods[0].signature.params 19:35
p6eval rakudo c24da0: OUTPUT«0␤0␤»
masak r: class A { method foo() {} }; say +grep { .positional && !.invocant }, A.^methods[0].signature.params; class B { has A $!a handles <foo> }; say +grep { .positional && !.invocant }, B.^methods[0].signature.params
p6eval rakudo c24da0: OUTPUT«0␤1␤»
masak jnthn: why the 1?
this surprised me greatly. 19:36
r: class A { method foo() {} }; class B { has A $!a handles <foo> }; say (grep { .positional && !.invocant }, B.^methods[0].signature.params).name
p6eval rakudo c24da0: OUTPUT«No such method 'name' for invocant of type 'List'␤ in block <anon> at /tmp/fUdadEjW3e:1␤␤»
jnthn masak: Maybe handles installs a closure rather than an anonymous method or something.
masak r: class A { method foo() {} }; class B { has A $!a handles <foo> }; say (grep { .positional && !.invocant }, B.^methods[0].signature.params)[0].name
p6eval rakudo c24da0: OUTPUT«$c␤»
masak what's $c?
masak jnthn: yeah, but it only does this when the attribute is private. 19:37
suddenly my nice introspection breaks :)
jnthn ...what?
It's about handles. It's nothing to do with privacy.
masak well, it shouldn't have anything to do with privacy. 19:38
masak but note that the output is different if it's $.a or $!a above. 19:38
that's what's surprising, more than anything else, really.
masak submits rakudobug 19:39
jnthn ah, maybe it goes thorugh a differnet codepath or something becuase you want it to call the accessor in the first case
I dunno
masak the B.foo method comes back saying it has one positional non-invocant parameter. 19:40
jnthn Well, did you check what kind of object it is?
It's probably a Block rather than a Method.
masak r: class A { method foo() { say "OH HAI" } }; class B { has A $!a handles <foo> }; B.new.foo
p6eval rakudo c24da0: OUTPUT«OH HAI␤»
masak r: class A { method foo() {} }; class B { has A $!a handles <foo> }; say (grep { .positional && !.invocant }, B.^methods[0].signature.params)[0].^name 19:41
p6eval rakudo c24da0: OUTPUT«Parameter␤»
masak er.
r: class A { method foo() {} }; class B { has A $!a handles <foo> }; say B.^methods[0].^name
p6eval rakudo c24da0: OUTPUT«Method␤» 19:42
masak it's a Method.
with a $c parameter.
jnthn Oh.
oh, of course.
it's |$c
As in, the capture
I'm surprised that is counted as positional. 19:43
I suspect that's the real bug.
r: sub foo(|$c) { }; say &foo.params[0].positional
p6eval rakudo c24da0: OUTPUT«No such method 'params' for invocant of type 'Sub'␤ in block <anon> at /tmp/GsWgBwfdNM:1␤␤»
jnthn r: sub foo(|$c) { }; say &foo.signature.params[0].positional
p6eval rakudo c24da0: OUTPUT«True␤»
pmichaud r: sub foo(|$c) { }; say &foo.count; 19:44
jnthn Yeah, that probably should not be True
p6eval rakudo c24da0: OUTPUT«Inf␤»
pmichaud r: sub foo(|$c) { }; say &foo.arity;
p6eval rakudo c24da0: OUTPUT«0␤»
pmichaud at least those are correct. :-)
masak I think the real bug is that the actual parameters in the handled method get hidden in a capture parameter -- and that this only happens for private attributes :/
jnthn What "actual parameters"?
masak jnthn: in this case, an empty list of them.
jnthn masak: Your expectations are weird. 19:45
masak: How is it meant to know that the object that you call .foo on will take no parameters?
pmichaud in some sense, wrappers that pass-through their arguments to the thing they wrap perhaps ought to expose the underlying signature, too. 19:46
jnthn IT'S A METHOD DISPATCH. WE DO NOT KNOW.
masak jnthn: what pmichaud said.
jnthn: why does it work for the public case but not for the private case?
pmichaud although yes, I see what jnthn++ is saying also, here.
masak well, me too. 19:47
jnthn masak: Because you're looking at the wrong method, perhaps?
.methods[0]
You're probably looking at the generated accessor method there.
pmichaud there have been a few times I've wished for .methods.<xyz> :-) 19:48
jnthn Sadly, it's a list, not a hash. 19:49
Wait...what? :)
jnthn .find_method('xyz') # :) 19:49
pmichaud ...where .postcircumfix:<{ }> greps through that list for method names
I can live with find_method
r: say 3.find_method('Str') 19:50
p6eval rakudo c24da0: OUTPUT«No such method 'find_method' for invocant of type 'Int'␤ in block <anon> at /tmp/kwRwcKAoyO:1␤␤»
jnthn .^
pmichaud r: say 3.^find_method('Str')
jnthn It's a meta-method
p6eval rakudo c24da0: OUTPUT«Str␤»
pmichaud r: say 3.^find_method('Str').signature
p6eval rakudo c24da0: OUTPUT«:(Mu , , Mu *%_)␤»
jnthn Um. 19:51
pmichaud okay, that works well enough for me I guess. :)
moritz you can also do .can('Str')[0]
jnthn Aside from Signature.perl needs some love.
pmichaud r: class A { method foo() { say "OH HAI" } }; class B { has A $!a handles <foo>; B.new.^find_method('foo').say
p6eval rakudo c24da0: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 2, near ""␤»
pmichaud r: class A { method foo() { say "OH HAI" } }; class B { has A $!a handles <foo>; }; B.new.^find_method('foo').say 19:52
p6eval rakudo c24da0: OUTPUT«foo␤»
pmichaud r: class A { method foo() { say "OH HAI" } }; class B { has A $!a handles <foo>; }; B.new.^find_method('foo').signature.say
p6eval rakudo c24da0: OUTPUT«:(Mu , |$c, Mu *%_)␤»
pmichaud there ya go. :-)
jnthn Yeah, it knows its a capture.
I bet that way, changing it to $.a doesn't change things.
pmichaud and we can't know anything about the underlying A.foo because it's a dispatcher, not a wrapper? 19:53
jnthn What is I wrote a class C is A { method foo($x) { } }
*if
It's a dynamic (at runtime) dispatch.
pmichaud exactly. the handles<foo> doesn't wrap the A::foo method. 19:54
It simply takes anything coming in at B.foo and sends it to $!a
jnthn Exactly. It installs a new method in B.
pmichaud wfm
masak ok, so "it's like this". 19:55
moritz ... "and that's the way it is"
masak r: class A { method foo() { say "OH HAI" } }; class B { has A $.a handles <foo>; }; say B.new.^find_method('foo').signature
p6eval rakudo c24da0: OUTPUT«:(Mu , |$c, Mu *%_)␤»
jnthn So, the only bug I see in all of this is that a |$c gives .positional as True under introspection.
pmichaud to fix the .positional problem, I propose adding $SIG_ELEMENT_IS_CAPTURE to
($!flags +& ($SIG_ELEM_SLURPY_POS +| $SIG_ELEM_SLURPY_NAMED)) == 0 &&
jnthn pmichaud: Yes, that should do it. 19:56
masak jnthn: it also counts |$c differently for the public/private attr cases.
even though it's there in both cases.
jnthn masak: You haven't shown that.
masak: [0] in the list of methods is probably not the generated 'foo', but instead the generated accessor method. 19:57
masak oh. 19:58
jnthn Relying on ordering of generated stuff in the method list is not wise. 19:58
pmichaud ...which is why I'd like a .^methods{'foo'} :-)
pmichaud I agree that find_method works... but it has that "find_" part in it. 19:58
masak r: class A { method foo() {} }; class B { has A $.a handles <foo> }; say (grep { .positional && !.invocant }, B.^methods[0].signature.params)[0].name 19:59
p6eval rakudo c24da0: OUTPUT«No such method 'name' for invocant of type 'Nil'␤ in block <anon> at /tmp/tqiggdrcJ4:1␤␤»
masak r: class A { method foo() {} }; class B { has A $.a handles <foo> }; say B.^methods[0].name
p6eval rakudo c24da0: OUTPUT«a␤»
pmichaud ding!
masak jnthn++ masak--
shoulda checked ;) 20:00
jnthn pmichaud: .^method_table{'foo'} may well work. 20:01
pmichaud: But it's...ugly again probably ;)
pmichaud jnthn: no, I like that one.
jnthn pmichaud: Be aware that method_table is only the methods *directly* in the class.
pmichaud find_method is too.... active for me
oh, that's a very good point.
jnthn pmichaud: Well, find_method can be very active :)
pmichaud Okay, I accept find_method.
jnthn pmichaud: If there's any fallback stuff or wildcard delegation going on, find_method is gonna also get you that too. 20:02
pmichaud ...we have no tests for .positional ?
jnthn .^methods is purely asking the class about what was added to it, and returns everything up to Cool or Any.
masak pmichaud: find_method is definitely active in the case of, say, Grammar::Debugger ;)
jnthn Right, that's an excellent example.
pmichaud: S06-signature/introspection.t will contain them if they're anywhere 20:04
pmichaud jnthn: it does not. I'll add a couple :)
jnthn :)
pmichaud r: sub xyz(|$c) { }; say &xyz.signature.params[0].name 20:05
p6eval rakudo c24da0: OUTPUT«$c␤»
pmichaud r: sub xyz(|$c) { }; say &xyz.signature.params[0].positional 20:06
p6eval rakudo c24da0: OUTPUT«True␤»
jnthn That's the one that I think should be False.
ooc
r: sub xyz(*@a) { }; say &xyz.signature.params[0].positional
p6eval rakudo c24da0: OUTPUT«False␤»
pmichaud that's explicitly checked for by SIG_ELEM_SLURPY_POS
jnthn ah, yes 20:07
pmichaud 19:55 <pmichaud> ($!flags +& ($SIG_ELEM_SLURPY_POS +| $SIG_ELEM_SLURPY_NAMED)) == 0 &&
r: sub xyz(|$c) { }; say &xyz.signature.params[0].slurpy 20:08
p6eval rakudo c24da0: OUTPUT«False␤»
pmichaud ...wonder if that should be true :-) 20:09
jnthn Arguably
pmichaud > sub xyz(|$c) { }; say &xyz.signature.params[0].positional 20:10
False
jnthn (The reason there's something to argue mostly being because of its snapshotting semantics.)
pmichaud spectesting 20:11
moritz pmichaud: is &xyz.count still Inf? 20:13
pmichaud > say &xyz.count 20:17
Inf
dalek kudo/nom: 3e00089 | pmichaud++ | src/core/Parameter.pm:
Fix .positional for capture parameters in signatures. masak++
 20:18
ast: 7a77ff8 | pmichaud++ | S06-signature/introspection.t:
[S06]: Add tests for capture parameter introspection of .positional, .named, etc.
masak lol I bloggied! strangelyconsistent.org/blog/july-7...-restoring 20:27
I again invite people to break the game in ways I didn't predict.
tadzik _sri: yeah, this sucks a bit but it's the only way I know of 20:38
there's also Select PMC in Parrot, and IO::Select is a wrapper around that, but it behaved a bit weird to me last time I tried it
_sri tadzik: *a bit* is quite an understatement ;p
tadzik :) 20:42
I was secretly hoping that it'd work fine for you
I wrote tests for it, but it was a bit like "hmm, what makes sense _and_ works..."
not sure if I ever succeeded on the first part 20:44
I have some free time in the upcoming days, maybe I could write some poll/epoll on parrot/nqp level which isn't completely stupid
dalek p: 8062b32 | pmichaud++ | src/QRegex/Cursor.nqp:
[qregex] Add !cursor_more, a preliminary method for handling :g, :ov, :ex.
 20:51
p: 052b7a1 | pmichaud++ | src/pmc/qrpa.pmc:
[qrpa] Fix exception message for shift from empty array (jnthn++).
masak good night, #perl6 21:22
sorear how much interest is there in getting github issues on dalek? 22:03
diakopter the command to close an issue could be 'exterminate' 22:04
sergot good night o/ 22:11
pmichaud github issues on dalek +1 22:13
jnthn 'night, #perl6 22:17
seldon Are the d, e, f, l, r, s, t, w, x, and z methods of class IO part of the public API? 22:21
sorear Yes 22:25
seldon What do they do? 22:28
pmichaud they're filetest operators 22:29
.d == is this file a directory
.e == does this file exist
seldon Oh, I see.
pmichaud it's what enables things like "foo.txt".io ~~ :f # true if foo.txt exists and is readable 22:30
see "Other things" in perlcabal.org/syn/S32/IO.html
dalek p: 7995664 | pmichaud++ | src/ops/nqp.ops:
Clean up some gcc compiler warnings.
 22:31
p: eb7f09e | pmichaud++ | src/6model/ (4 files):
More compiler warning cleanups.
p: 28f89ef | pmichaud++ | src/6model/base64.c:
More casting and code cleanup.
seldon There they are. I didn't make the connection to member functions :o 22:35
I'm shamelessly lifting method signatures from IO.pm, btw. I hope that's okay. 22:37
diakopter pmichaud: does nqp's --target=past still work? It tells me Could not find sub _dumper 23:24

Powered by Cro and the Raku® Programming Language.

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

Thank you!