🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). Log available at irclogs.raku.org/raku/live.html . If you're a beginner, you can also check out the #raku-beginner channel!
Set by lizmat on 6 September 2022.
SmokeMachine m: use experimental :rakuast; say RakuAST::VarDeclaration ~~ Any 06:05
camelia ===SORRY!===
Cannot find method 'archetypes' on object of type (null)
SmokeMachine m: use experimental :rakuast; say RakuAST::VarDeclaration.^mro 06:06
camelia Method KnowHOW.mro not found
in block <unit> at <tmp> line 1
SmokeMachine github.com/rakudo/rakudo/issues/5678 06:13
SmokeMachine now it's a bit easier to understand what ASTQuery is doing: github.com/FCO/ASTQuery?tab=readme...file#debug 06:31
timo i'm not sure i can make the knohows turn into fully realized objects with metaobjects and whatnot, but i could probably build some nqp::syscall(blah, $theobj) that let you introspect knowhow-repred objects 10:00
lizmat the problem was that RakuAST::VarDeclaration was a package, not a class 10:24
lizmat weekly: github.com/renormalist/raku-tools-...completion 11:45
notable6 lizmat, Noted! (weekly)
SmokeMachine m: use experimental :rakuast; say RakuAST::VarDeclaration::Simple ~~ Any 12:23
camelia True
SmokeMachine Sorry, my fault 12:26
lizmat what is? 12:30
m: use experimental :rakuast; say RakuAST::VarDeclaration ~~ Any
camelia True
SmokeMachine The VarDeclaration confusion…
Oh! I misunderstood your message then! 12:31
lizmat m: class A::B { }; dd A.^mro 12:33
camelia No such method 'mro' for invocant of type
'Perl6::Metamodel::PackageHOW'
in block <unit> at <tmp> line 1
lizmat which is a correct message I think
the other one was caused by some bootstrap issues, which I worked around by making it a class
SmokeMachine Yes, makes sense. Thanks 12:34
lizmat yw 12:39
afk& 12:58
SmokeMachine lizmat: have I told you about this issue? github.com/lizmat/rak/issues/5 18:32
I’m not sure if that should be there or in other repo…
tbrowder ok, got 'rak' doing what i want on debian. i used this command: "rak --find --is-readable --paths=/usr/share/fonts,/User,~/Library/Fonts --extensions=pfb,woff,otf,ttf --absolute" 20:03
that should also work on mac. i would appreciate a mac user trying it and see if it finds any fonts on your host 20:04
the advantage to me is is makes using a Build script easier on multiple OSs
[Coke] I think you might mean /Users 20:15
tbrowder ok, i’ll change that. thanks! 20:25
did it find any fonts? 20:27
[Coke] ** Stopped showing results after 1000 matches ** 20:34
SmokeMachine tbrowder: It returned many lines here… 20:35
(with /Users) 20:36
tbrowder SmokeMachine: thank you! and thanks to lizmat for an awesome multi-os tool! 20:42
talk about a giant Swiss raku knife…what a swell Christmas present 20:44
patrickb I have the following token in a grammar: `token routine { <!ww> @routines <!ww> }` with @routines having 960 entries. That token is really slow. Of the 70 seconds it takes to parse 144 lines of input more than the half is this token alone. Is this a stupid way of searching for a fixed set of tokens? How could I do better? 22:06
[Coke] is <?wb> faster than <!ww> ? 22:14
and or do you want << @routines >> ? 22:15
wondering if either of those are faster
guifa I think technically when you do @routines it treats it as
@routine[0] | @routine[1] | @routine[2] |... 22:16
so then LTM kicks in so it has to check all of them
patrickb guifa: How would I get around that? 22:21
I don't need LTM here.
ab5tract There should be some way to deactivate that… 23:02
lizmat {} 23:12
ab5tract patrickb: It’s not exactly elegant, but maybe using @routines.join(“ || “) in an EVAL’d regex might work?
lizmat afaik, an empty code block in a regex will stop LTM 23:13
afk again 23:15
ab5tract That seems kind of random :)
tbrowder and [Coke] also thanks for checking
ab5tract Feels like an adverb would be more discoverable 23:16