01:09 stanrifkin left 01:18 Manifest0 left 01:36 librasteve_ left 04:53 arkiuat left 05:06 arkiuat joined 05:11 arkiuat left 05:18 kjp left, kjp joined 05:24 deoac left 05:39 arkiuat joined 05:44 arkiuat left 05:56 arkiuat joined 06:01 arkiuat left 06:14 arkiuat joined 06:18 arkiuat left 06:30 arkiuat joined 06:39 arkiuat left 06:51 arkiuat joined 06:55 arkiuat left 07:09 arkiuat joined 07:13 arkiuat left 07:25 arkiuat joined 08:29 arkiuat left 08:35 arkiuat joined 08:40 arkiuat left 09:11 arkiuat joined 09:16 arkiuat left, jmcgnh left 09:17 jmcgnh joined 09:36 arkiuat joined 09:40 arkiuat left 10:09 arkiuat joined 10:14 arkiuat left 10:24 Manifest0 joined 10:37 arkiuat joined 10:41 arkiuat left 11:09 arkiuat joined 11:14 arkiuat left 11:29 arkiuat joined 11:42 arkiuat left 11:54 arkiuat joined 11:58 arkiuat left 12:20 arkiuat joined 12:24 arkiuat left 12:56 arkiuat joined 13:01 arkiuat left 13:24 arkiuat joined 15:33 arkiuat left 15:45 arkiuat joined 15:48 Manifest0 left 15:50 arkiuat left 15:55 arkiuat joined 17:29 librasteve_ joined, human_blip left 17:32 human_blip joined 19:40 arkiuat left 19:53 arkiuat joined 19:57 arkiuat left 20:09 arkiuat joined
disbot4 <Welemtam> Hi! I'm learning OOP in Raku. I got confused by one thing. I defined : class Human { has $!name; has $!age is required; has $!nationality; method new($name,$age,$nationality) { self.bless(:$name,:$age,:$nationality); } method is-adult { $!age >= 18; } } From the Raku guide, overriding the default new method like above allows to pass non-named args (why not). At first, is-adult was complaining that 20:31
$!age did not always have a value, so I put "is required". But now, I have : "The attribute '$!age' is required, but you did not provide a value for it.". I guess it's because of my override of "new", but is there a way to get around that error ? Thanks!
<librasteve> please can you share the code you have to make a new instance? 20:36
<Welemtam> Sure : my $john = Human.new('John', 95, 'American'); It uses my new constructor , nothing comes undefined from the caller at least 20:37
<librasteve> tx 20:38
<librasteve> looks like a bug to me ... please can you file an issue? 20:41
<librasteve> (maybe I will be corrected...)
lizmat Welemtam not a bug: is there a reason you used the ! twigil on the attributes ? 20:49
it will work if you specify the attributes as: has $.name; has $.age; has $.nationality 20:50
disbot4 <Welemtam> To make them private (reading the raku guide at raku.guide/#_encapsulation)
lizmat but you still want to be able to specify them in a .bless, right ? 20:51
also: do you want to be able to do $john.age to find out the age of $john ?
disbot4 <Welemtam> Well the guide was suggesting to do that in order to define a constructor taking non-named parameters. But maybe that's already a weird idea. 20:52
lizmat no, it's not... it's just that it also affects what .bless accepts
disbot4 <Welemtam> I could imagine that I may want $!age to stay private for encapsulation matters, and only give access to the is-adult method. 20:53
lizmat where did you find the reference to make attributes private that way ?
disbot4 <librasteve> .oO bless args are constrained by the attr definitions ... sorry my bad
lizmat librasteve .new is reallly just thin shell around .bless :-)
Welemtam in that case: has $!age is built 20:54
disbot4 <Welemtam> In the link above (raku guide), §9.2 defines a class Human, stating that "! is used to explicitly declare that the attribute is private"
lizmat that will *not* create the accessor, but it *will* allow you to specify the age with .bless
ok, we will need to fix that example 20:55
disbot4 <Welemtam> Thanks!
lizmat yw 20:56
&
disbot4 <Welemtam> Actually the example by itself in the guide is not using bless on private attributes. I did merge §9.2 and §9.3 samples when playing around "naively" 🙂 21:00
21:08 deoac joined
disbot4 <nahita3882> yeah you cannot bless private attributes 21:09
<nahita3882> therefore $!attr is required; can never work, I think 21:10
<nahita3882> therefore the error message could be improved
<nahita3882> also if I may rant, the methods taking whatever named args with no issue is annoying
<nahita3882> here self.bless takes :name etc. but doesn't use them at all
<nahita3882> can do self.bless(:asdasdasdasdas) and it's as okay as it stands now 21:11
21:19 msiism joined
msiism If I have a string of the form 'key: value', what would be an efficient way to extract just the value? 21:24
I mean something like split on ':', and giev me the right side, trimmed of leading and trailing whitespace. 21:25
disbot4 <librasteve> m: 'key: value'.split(': ')[1].say
<Raku eval> value 21:26
msiism Oh, okay. 21:27
disbot4 <librasteve> m: 'key: value'.split(':')[1].trim.say 21:28
<Raku eval> value
<librasteve> (does both leading and trailing)
msiism So, I was just confused about Seqs then, it seems. 21:29
I tought you couldn't access individual elements of a Seq by index.
21:30 arkiuat left
msiism So, is the main difference between Seqs and arrays that Seqs can't be sparse then? 21:31
And are normally consumed during iteration…
disbot4 <librasteve> docs.raku.org/type/Iterable#typegraphrelations 21:34
<librasteve> in a nutshell, Seq does Iterable, List also does Positional and Array is a mutable List 21:38
msiism Interesting. 21:39
disbot4 <librasteve> in raku very many functions return a Seq (ie a lazy, iterate-once list) than can then be exhausted ... assigning to List and Array basically eagerly consumes the Seq 21:41
msiism But, if Seq does not implement Positional, why can you access elements of a Seq by index? 21:42
21:54 arkiuat joined
msiism Okay, "the Seq class does provide some positional subscripting" (docs.raku.org/language/list). 21:56
21:58 arkiuat left 22:24 arkiuat joined
lizmat Seq does the PositionalBindFailover role 22:27
22:28 arkiuat left 22:45 msiism left 22:58 arkiuat joined 23:05 arkiuat left 23:27 arkiuat joined 23:32 arkiuat left 23:42 deoac left 23:47 arkiuat joined 23:55 arkiuat left

Powered by Cro and the Raku® Programming Language.

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

Thank you!