This channel is intended for people just starting with the Raku Programming Language ( Logs are available at
Set by lizmat on 8 June 2022.
00:25 ab5tract left 00:42 QhpAptyj9hj0RQwM left 03:34 rf left 05:35 Heptite left 05:54 kjp left 05:56 kjp joined 09:03 frost joined 09:09 dakkar joined 09:11 ab5tract joined 09:14 frost left 09:18 frost joined 09:19 frost left 10:25 QhpAptyj9hj0RQwM joined 10:31 ab5tract left 10:48 ab5tract joined 11:09 ab5tract left 11:11 QhpAptyj9hj0RQwM left 11:28 QhpAptyj9hj0RQwM joined 11:33 QhpAptyj9hj0RQwM left
Nahita is this expected 12:15
m: .^name.put if (True given 4) # puts Any
Raku eval Any
Nahita $_ is not transported to the left side 12:16
12:20 ab5tract joined
lizmat I think so, as here will be an implicit scope with the (True given 4) 12:22
Nahita hmm, so parantheses are not only for precedence adjustment but also introduce scope? 12:24
i can remove the parantheses in this case 12:25
12:25 frost joined
lizmat no, generally () only group 12:25
12:26 frost left, frost joined
Nahita m: die "this is bad $_" if 2 !~~ (-$_ ..^ $_) given 3 12:30
Raku eval Exit code: 1 this is bad 3 in block <unit> at main.raku line 1
Nahita shouldn't this do nothing?
it errors with "this is bad 3" 12:31
m: say 2 !~~ (-3 ..^ 3)
Raku eval False
Nahita this says False
which is expected as 2 is in that Range 12:32
12:33 ab5tract left
OH 12:33
smartmatch changes the $_ there...
going to use chained comparison instead... 12:34
lizmat yeah, that feels like a good idea :-)
12:35 frost left
Nahita it works thanks 12:36
wouldn't it be cool if the standard Test module said "all pass" or something if all tests passed 12:37
at the end
it makes me hesitate 12:38
lizmat it complains at the end if tests failed 12:40
if it doesn't complain, all tests expected to pass, did pass 12:41
12:43 frost joined 12:46 frost left
Nahita yes, i thought it'd be better if it said something like "all tests passed" but maybe not 12:47
Nemokosch smartmatch to expressions with $_ is a really funky trap 12:53
I think that one is mentioned in the docs
12:53 frost joined 12:54
13:03 ab5tract joined 13:19 ab5tract left 13:29 ab5tract joined 13:35 ab5tract left 13:36 ab5tract joined
el gatito (** advocate) tbh $_ should only be used as an “implicit argument variable” and nothing else 13:51
also the exception handling pattern in raku is like bruhperl try { # do somethin’ } if $! { say ‘error’; } 14:00
lizmat in raku, the pattern is: 14:03
say "error"
try is used in case you're *NOT* interested in why something failed 14:04
14:04 frost left 14:07 ab5tract left
el gatito (** advocate) ok its actually just my irrational anger towards raku’s nonconformist syntax if i do this i should be gtgperl try { die ‘i died’; CATCH { when X::AdHoc { say ‘an ad hoc error’ } } } 14:11
lizmat if you do a .resume in the when { } block, you wouldn't need the try 14:16
Nemokosch try is not even an essentially part of that structure
lizmat m: CATCH { say "died"; .resume }; die "now"; say "still alive"
camelia died
still alive
Nemokosch also, Raku doesn't feed you with unnecessary exceptions for anything nonconventional - which is the right choice in the 21st century, I'd say 14:19
perhaps this is the worst design decision in Python I can think of - raising exceptions, ranging from a missing key to ending the iteration for an iterator 14:20
14:21 QhpAptyj9hj0RQwM joined
$! seems like a nice and convenient shortcut to avoid actual exception handling, imo 14:22
14:37 jgaz joined 15:10 Heptite joined 15:14 QhpAptyj9hj0RQwM left
el gatito (** advocate) KeyError is fine imo, you expect the key to exist when indexing but it doesn’t, so it throws an exception and aren’t raku control exceptions a more general version of python’s StopIteration? 15:27
Nemokosch I don't think KeyError is fine - one would almost never want to escalate a missing key as an error 15:30
especially considering that once it's an exception, it may be caught at a completely different place, and that's really only up to the author to design proper checks for what exceptions they handle and what they escalate further 15:31
also, one rarely ever needs to do anything about control exceptions (I'm kinda salty warn produces a control exception) 15:36
to have such a shitty interface for iterators, though, is unforgivable, lol 15:37
the Raku interface is still not great but yeah, way better than an interface that downright relies on exceptions... 15:38
el gatito (** advocate) you are saying python?
Nemokosch basically answered by now 15:39
el gatito (** advocate) py iterators should use iterator.done instead of throwing an exception 15:41
also python’s overzealousness in throwing exceptions might have something to do with it being a beginner language? like its better to shout at them when their code does something wrong than to ignore it and cause problems later on 15:46
Nemokosch could be, I don't know 15:47
it's really the worst iterator interface I can quickly think of
el gatito (** advocate) 😔 15:51
frost m: dd **.Int 15:53
Raku eval sub (*@_) { #`(Sub|5526418166944) ... }
frost m: (**.Int)(1,2,3) 15:54
Raku eval
frost m: (**.Int)(1,2,3).say
Raku eval (1 2 3)
frost I don't understand this code **.&foo, what's going on 😟 15:56
el gatito (** advocate) about iterators php actually has the most complicated and flexible one
Nemokosch where did you find it? 15:57
el gatito (** advocate) ** is a HyperWhatever
frost Why I don't find it in docs
el gatito (** advocate) and it turns the expression around it into a callable
Nemokosch PHP has been built into a huge language
To be honest - it's a better question why you came across this 15:58
frost m: dd **
Raku eval **
Nemokosch like yes, it is in the docs but why are you bringing this all of a sudden?
I have used it literally once
and even then more for the lolz
el gatito (** advocate) you might start with normal whatevers @frost
Nemokosch as a sort of sentinel value, not to curry
el gatito (** advocate) m: say(*(3)) 15:59
Raku eval Exit code: 1 No such method 'CALL-ME' for invocant of type 'Whatever' in block <unit> at main.raku line 1
el gatito (** advocate) m: say((*)(3))
Raku eval Exit code: 1 No such method 'CALL-ME' for invocant of type 'Whatever' in block <unit> at main.raku line 1
frost Actually I find it from the weekly news
el gatito (** advocate) m: say((*+1)(3))
Raku eval 4
frost @el gatito (** advocate) I know whatevercode
But not familiar with hyperwhatever 16:00
Nemokosch what code example did you see?
el gatito (** advocate) and i don’t know what hyperwhatever does too so sorry 😭
frost @Nemokosch 16:02
See the raku solution 16:03
The o infix is also new to me...
Nemokosch I swear this is gonna be another raiph show-off 16:04
lizmat hehe
frost thanks lizmat, I found it but not found**
Nemokosch 16:05
frost Use the research bar I didn't find the doc of **
Only the infix that 1**2
lizmat 16:06
frost @Nemokosch thanks
el gatito (** advocate) its codegolf what do you expect 16:07
Nemokosch I revoke my original statement 16:08
raiph would never do something this mad
16:09 QhpAptyj9hj0RQwM joined
anyway, the way I read these hieroglyphs 16:16
> ($~+)o(**.FatRat) 16:17
to be honest, I'm not sure how the precedence adds up still 16:18
$~ is surely for string coercion, a dirty hack using the anonymous state variable
$ ~ something 16:19
frost m: $1
Raku eval
Nemokosch now, if $ is not set
frost m: $1.say
Raku eval Nil
Nemokosch it will throw a warning AND coerce to the empty string
16:19 QhpAptyj9hj0RQwM left
frost m: $~1.say 16:20
Raku eval 1 WARNINGS for /home/glot/main.raku: Useless use of "~" in expression "$~1.say" in sink context (line 1) Use of uninitialized value element of type Any in string context. Methods .^name, .raku, .gist, or .say can be used to stringify it to something meaningful. in block <unit> at main.raku line 1
16:20 QhpAptyj9hj0RQwM joined
Nemokosch m: dd $~1 16:20
Raku eval Use of uninitialized value element of type Any in string context. Methods .^name, .raku, .gist, or .say can be used to stringify it to something meaningful. in block <unit> at main.raku line 1 "1"
Nemokosch this is a string
frost m: ~1.say
Raku eval 1 WARNINGS for /home/glot/main.raku: Useless use of "~" in expression "~1.say" in sink context (line 1)
Nemokosch m: dd ~1
Raku eval "1"
Nemokosch this is also a string... however, with a different operator
that happens to have higher precedence 16:21
so using $ there is a precedence hack clearly
frost Yeah, why there need a $
Nemokosch it's cheaper than the parens 😆
frost m: say (~+o**.FatRat)("1","2") 16:25
Raku eval 3
el gatito (** advocate) m: my &sum = **.sum; say sum(1, 2, 3, 4);
frost say (~ * + *o**.FatRat)("1","2")
Raku eval (1 2 3 4)
el gatito (** advocate) what
frost m: say (~+o**.FatRat)("1","2") 16:26
Raku eval 3
el gatito (** advocate) m: my &sum = **.reduce({ $^x + $^y }); say sum(1, 2, 3, 4); 16:27
Raku eval (1 2 3 4)
el gatito (** advocate) m: my &add = ** + 1; say sum(1, 2, 3, 4); 16:28
Raku eval 10
el gatito (** advocate) m: my &add = ** + 1; say add(1, 2, 3, 4);
Raku eval (2 3 4 5)
el gatito (** advocate) o 16:29
Nemokosch m: my &add = ** + 1; dd add(1, 2, 3, 4); 16:30
Raku eval (2, 3, 4, 5).Seq
Nemokosch now this is interesting 16:31
I mean, apparently it performs a map
however, not the same map that the hyper metaoperator would
just an "oldschool" map
m: dd (1, 2, 3, 4) >>+>> 1 16:32
Raku eval (2, 3, 4, 5)
Nemokosch this is a List, not a Seq
el gatito (** advocate) what is the difference between list and seq 16:36
Nemokosch the interface 16:39
16:48 QhpAptyj9hj0RQwM left 16:51 QhpAptyj9hj0RQwM joined
el gatito (** advocate) show me a concrete example 16:58
Nemokosch could you not just look up the docs? 17:01
17:10 QhpAptyj9hj0RQwM left
el gatito (** advocate) sure 😤 17:39
17:39 dakkar left 17:43 QhpAptyj9hj0RQwM joined
well they are actually pretty different 17:44
Seq doesn’t have kv for example 17:45
lizmat m: say (^10).map(*+1).kv # I think it does? 17:46
camelia (0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10)
lizmat m: say (^10).map(*+1).^name 17:47
camelia Seq
Nemokosch Sadly, kv exists on Any 17:55
therefore, Seq, among other types, will just invisibly upgrade to List in order to serve you with kv 17:57
|| ☕ || 17:58
anyway, a Seq isn't even Positional 17:59
what it has is a potential to upgrade to one (PositionalBindFailover)
lizmat ah hmmm... maybe that's an issue that there is no Seq.kv then
m: say (^10).map(*+1)[3] # PositionalBindFailover in action 18:00
camelia 4
Nemokosch m: .^methods>>.gist.grep('kv').&dd given (^10).map(* + 1) 18:04
Raku eval ().Seq
Nemokosch m: .^methods>>.gist.grep('kv').&dd given (^10).map(* + 1).list
Raku eval ("kv",).Seq
Nemokosch the difference
lizmat Seq should probably have a .kv method 18:05
20:46 ab5tract joined 21:10 QhpAptyj9hj0RQwM left 21:39 QhpAptyj9hj0RQwM joined 22:08 jgaz left
p6steve nah 22:34
22:47 QhpAptyj9hj0RQwM left
Nemokosch well it depends on the implementation of Seq.list 22:53
I guess it's supposed to be a performance change rather than a public interface change
i.e why fetch the known universe when at the end of the day you can be as lazy as with the original sequence 22:54
23:56 ab5tract left