🦋 Welcome to Raku! raku.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: colabti.org/irclogger/irclogger_log/raku
Set by ChanServ on 14 October 2019.
SmokeMachine xinming_: www.irccloud.com/pastebin/VG7hKIbp/ 00:06
xinming_: it's working, but gives an error if you not pass a Ble to your Ble... 00:08
xinming_: it's working, but gives an error if you not pass a Ble to your Bla...
xinming_: the important part is: `has UInt $!ble-id is column{ :references{ Ble.id }, :!nullable }` 00:14
SmokeMachine red: model Bla { ... }; model Ble { has $!id is serial; has Bla @.blas is relationship{ .ble-id } }; model Bla { has $!id is serial; has UInt $!ble-id is column{ :references{ Ble.id }, :!nullable }; has Str $.name is column; has Ble $.ble is relationship{ .ble-id } }; my $*RED-DB = database "SQLite"; Bla.^create-table; Ble.^create-table; Ble.^create: :blas[{:name<bla>}, {:name<ble>}]; .say for Bla.^all; Bla.^create: 00:39
:name<bli>
SmokeMachine red: model Bla { ... }; model Ble { has $!id is serial; has Bla @.blas is relationship{ .ble-id } }; model Bla { has $!id is serial; has UInt $!ble-id is column{ :references{ Ble.id }, :!nullable }; has Str $.name is column; has Ble $.ble is relationship{ .ble-id } }; my $*RED-DB = database "SQLite"; Bla.^create-table; Ble.^create-table; Ble.^create: :blas[{:name<bla>}, {:name<ble>}]; .say for Bla.^all; Bla.^create: 00:40
:name<bli>
SmokeMachine red: model Bla { ... }; model Ble { has $!id is serial; has Bla @.blas is relationship{ .ble-id } }; model Bla { has $!id is serial; has UInt $!ble-id is column{ :references{ Ble.id }, :!nullable }; has Str $.name is column; has Ble $.ble is relationship{ .ble-id } }; my $*RED-DB = database "SQLite"; Bla.^create-table; Ble.^create-table; Ble.^create: :blas[{:name<bla>}, {:name<ble>}]; .say for Bla.^all; Bla.^create: 00:43
:name<bli>
Geth problem-solving/old-issue-tracker: 57f7054251 | (Aleks-Daniel Jakimenko-Aleksejev)++ | solutions/rakudo/Old-Issue-Tracker-Dismantling-Plan.md
Mention others too
 02:06
problem-solving/old-issue-tracker: 57f2e8ebb8 | (Aleks-Daniel Jakimenko-Aleksejev)++ | solutions/rakudo/Old-Issue-Tracker-Dismantling-Plan.md
Mention that perl6/ will likely not be “perl6/” later
Geth ecosystem: 9b15b778ca | tony-o++ (committed using GitHub Web editor) | META.list
sqlite orm
 03:42
discord6 <Aearnus> does cro/cro::webapp have anything like partials 06:39
<Aearnus> a-la ruby on rails' <%= render ... %>
<Aearnus> I couldn't find anything like that online 06:40
<Aearnus> aha, cro::webapp provides <:use '...'>
<Aearnus> 😄 06:41
Geth doc: 3b828abad6 | (Stoned Elipot)++ | doc/Language/classtut.pod6
Pursue .WHAT removal

  - Sync output with code example
  - Update code example description
  - Advise .^name usage		 06:48
doc: 9fc140b745 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | doc/Language/classtut.pod6
Merge pull request #3074 from stoned/classtut-no-what

Pursue .WHAT removal We can discuss whether .WHAT should be removed from this particular document, but this PR just makes the text match the example.
tbrowder tony-o: hi, how do you create a "unique(field1, field2)" constraint in xoo? 11:24
xinming_ SmokeMachine: When a has_many relationship is defined, How can we further filter the has_many? 11:26
For example, $object.has_many_items.^load: xxxx;
When I try this, I got ` No such method 'load' for invocant of type 'Perl6::Metamodel::ClassHOW'. Did you mean 'head'?` 11:27
SmokeMachine: Got it, It's a bit confusing that some api are not "unified", we need to use .grep instead of .^grep, `Model.^load: 1;` works, but `$xx-object.has_many_items.^load: 1;` doesn't. with has_many_items relationship, We need to use $xx-object.has_many_items.grep: ...; The API is bit confusing. 11:34
xinming_ SmokeMachine: Also, How do we do multi-key primary key? ;-) 11:48
SmokeMachine: Also, the way to get all columns from a row. 12:10
SmokeMachine xinming_: $obj.has-many.first: *.name eq “bla” 12:21
xinming_: you can have multiple `is id` fields... 12:26
xinming_: the obj has all columns... 12:27
SmokeMachine xinming_: `$obj.has-multi` returns an ResultSeq, so, you can do with it anything you would do with `Model.^all`... 13:12
SmokeMachine xinming_: have you seen my answers? 13:44
xinming_: colabti.org/irclogger/irclogger_lo...-10-26#l76 13:45
SmokeMachine .tell xinming_ colabti.org/irclogger/irclogger_lo...-10-26#l76 13:46
tellable6 SmokeMachine, I'll pass your message to xinming_
SmokeMachine red: model Bla { ... }; model Ble { has $!id is serial; has Bla @.blas is relationship{ .ble-id } }; model Bla { has $!id is serial; has UInt $!ble-id is column{ :references{ Ble.id }, :!nullable }; has Str $.name is column; has Ble $.ble is relationship{ .ble-id } }; my $*RED-DB = database "SQLite"; Bla.^create-table; Ble.^create-table; Ble.^create: :blas[{:name<bla>}, {:name<ble>}]; .say for Bla.^all; Bla.^create: 14:26
:name<bli>
xinming_ SmokeMachine: Thanks. 15:05
tellable6 2019-10-26T13:46:01Z #raku <SmokeMachine> xinming_ colabti.org/irclogger/irclogger_lo...-10-26#l76
xinming_ I'll reconnect, bll 15:06
SmokeMachine: What I mean is, wether there is a method to call, so we can get all columns data from model.
xinming back with new config, seems ssl on freenode will kill me for a period. 15:10
SmokeMachine xinming_: Model.^columns
xinming SmokeMachine: How about the object?
SmokeMachine xinming: $obj.^columns 15:11
xinming SmokeMachine: the ^columns returns an Hash with values? 15:12
SmokeMachine xinming: no, a list of `Attribute`s
ZzZombo Remind me if I can declare custom `has` traits? 15:13
SmokeMachine ZzZombo: sorry, what do you mean? 15:14
ZzZombo I know you can declare custom `is` traits, but back then couldn't `has`. Did that change? 15:15
SmokeMachine not that I know.. 15:17
red: say 42 15:19
vrurg ZzZombo: you can't declare a custom trait yet. Not sure if it ever be possible. 15:21
ZzZombo But why not? What's the difference between "is" and "has" that allows to declare the former and not the latter? 15:22
vrurg ZzZombo: the difference is in grammar. Traits are handled individually each one. 15:24
ZzZombo: actually, it should be possible to define a slang and add a custom trait. 15:25
The problem with traits is that they have different syntax. 15:26
vrurg is afk
SmokeMachine red: use Post; say Post.new.^columns # <- xinming 15:29
redable SmokeMachine, [Int $!id Int $!author-id Str $!title Str $!body Bool $!deleted DateTime $!created Set $!tags]␤
xinming red: use Post; say Post.new.^columns.map: { .perl.say }; 16:22
Why doesn't bot respond for me? 16:24
red: "hello".say;
timotimo y'know, we could totally have something short for the evalbot where it'd give you a bunch of models with example data to play around with 16:30
like "use RedEx" :D
and a specific output module that spits out CSV with headers, creating gists where the filename is (related to) the query code and the contents are the result as sortable/filterable tables 16:33
SmokeMachine xinming: I've turned it off... sorry 16:45
timotimo: it's a good idea... about the models, I've alread started: github.com/FCO/RedBot/tree/master/samples 16:47
SmokeMachine red: use Post; say Post.new.^columns.map: { .perl.say }; 16:56
redable SmokeMachine, post.id␤post.author_id␤post.title␤...tags␤(True True True True True True True)␤
pmurias hi 16:56
tellable6 2019-10-26T16:40:14Z #raku-dev <MasterDuke> pmurias: any idea why github.com/MasterDuke17/nqp/commit...fcfdd891d9 doesn't work and gives this? gist.github.com/MasterDuke17/14d6d...3d13d13bb8
MasterDuke hey hey 16:57
SmokeMachine red: use Post; say Post.^columns.map: { .column.name };
redable SmokeMachine, (id author_id title body deleted created tags)␤
SmokeMachine xinming: ^^ 16:58
pmurias: hi!
timotimo: any idea of models we should have? 17:07
timotimo no idea ... the typical ones i guess?!
todo application! states and their capitals around the world? 17:08
veesh a blog?
that's the other classic
uh.... ecommerce! 17:09
SmokeMachine red: say Post.^all.head.author # <- veesh
redable SmokeMachine, 04===SORRY!04=== Error while compiling /code/./code.p6␤Undeclared name:␤ Post used at line 2␤␤
SmokeMachine red: use Post; say Post.^all.head.author # <- veesh 17:10
redable SmokeMachine, (Person)␤
SmokeMachine red: use Person; Person.^create: :name<SmokeMachine>, :posts[{:title("new post"), :body("A new post")}] 17:11
redable SmokeMachine, gist.github.com/2fa3109ccbe52c8356...e9db16b74d
SmokeMachine red: use Post; say Post.^all.sort(*.id).tail.author # <- veesh
redable SmokeMachine, (Person)␤
SmokeMachine :( 17:12
red: use Post; .say for Post.^all
redable SmokeMachine, gist.github.com/b7b899950a0e06c5fc...175044de9d
xinming red: use Post; my Post $post .= new; $post does role { method columns-data () { my %colmap = self.^attr-to-column; %(self.^columns.map: { %colmap{.name} => .get_value(self) }); } }; $post.columns-data.perl.say; 17:17
redable xinming, {:author_id(Any), :body(Any), :created(DateTime.new(2019,10,26,17,17,34.04522466659546)), :!deleted, :id(Any), :tags(set()), :title(Any)}␤
xinming red: use Post; my Post $post .= new; $post does role { method columns-data () { %(self.^columns.map: { .column.name => .get_value(self) }); }; $post.columns-data.perl.say; 17:18
redable xinming, 04===SORRY!04=== Error while compiling /code/./code.p6␤Missing block␤at /code/./code.p6:3␤------> 03<BOL>08⏏04<EOL>␤
xinming red: use Post; my Post $post .= new; $post does role { method columns-data () { %(self.^columns.map: { .column.name => .get_value(self) }); } }; $post.columns-data.perl.say;
redable xinming, {:author_id(Any), :body(Any), :created(DateTime.new(2019,10,26,17,18,59.74856209754944)), :!deleted, :id(Any), :tags(set()), :title(Any)}␤ 17:19
xinming red: use Post; my Post $post = Post.^all.head; $post does role { method columns-data () { %(self.^columns.map: { .column.name => .get_value(self) }); } }; $post.columns-data.perl.say;
redable xinming, {:author_id(Any), :body(Any), :created(DateTime.new(2019,10,26,17,19,35.81536102294922)), :!deleted, :id(Any), :tags(set()), :title(Any)}␤
xinming SmokeMachine: I mean, the columns-data returns the column-value map from the db directly. 17:20
I tried to write the meta method in MetaModelX::Red::Model
Not enough understanding of it for now.
SmokeMachine red: use Post; my $post = Post.^all.head; $post.HOW does role :: { method column-values($obj --> Hash()) { $obj.^columns.map: { .column.name => .get_value: $obj } } }; say $post.column-values 17:26
redable SmokeMachine, No such method 'column-values' for invocant of type 'Post'␤ in block <unit> at ./code.p6 line 2␤␤
SmokeMachine red: use Post; my $post = Post.^all.head; $post.HOW does role :: { method column-values($obj --> Hash()) { $obj.^columns.map: { .column.name => .get_value: $obj } } }; say $post.^column-values
redable SmokeMachine, {author_id => (Any), body => body-test1, created => 2019-10-22T12:00:19.196347Z, deleted => 0, id => 1, tags => set(), title => test1}␤ 17:27
SmokeMachine xinming: ^^ 17:27
xinming SmokeMachine: I get that version too. 17:34
I think you can add this meta method.
It's useful.
xinming SmokeMachine: termbin.com/cwav 18:57
SmokeMachine: This test script will raise a bug. Maybe multi column relationship is not perfect.
SmokeMachine red: termbin.com/cwav 19:19
SmokeMachine xinming: a relationship should point to a foreign key, not to a primary key... 19:23
SmokeMachine red: termbin.com/cwav 19:25
redable SmokeMachine, Successfully fetched the code from the provided URL
SmokeMachine, 04===SORRY!04=== Error while compiling /code/./code.p6␤'EXPORTHOW::DECLARE::model' conflicts with an existing meta-object imported into this lexical scope␤at /code/./code.p6:6␤------> 03use Red08⏏04;␤
SmokeMachine www.irccloud.com/pastebin/JY6hC3p0/ 19:27
red: www.irccloud.com/pastebin/raw/JY6hC3p0/ 19:28
redable SmokeMachine, Successfully fetched the code from the provided URL
SmokeMachine, 04===SORRY!04=== Error while compiling /code/./code.p6␤'EXPORTHOW::DECLARE::model' conflicts with an existing meta-object imported into this lexical scope␤at /code/./code.p6:5␤------> 03use Red08⏏04;␤
SmokeMachine www.irccloud.com/pastebin/QUqMTYof/ 19:28
red: www.irccloud.com/pastebin/raw/QUqMTYof/ 19:29
redable SmokeMachine, Successfully fetched the code from the provided URL
SmokeMachine, gist.github.com/fe17b251363660ecf9...f9fe8b7990
SmokeMachine xinming: ^^
xinming SmokeMachine: Thanks, After adding the references, It worked.
SmokeMachine: yea, Your slogan is right for Red, It's WIP, but quite usable. 19:30
BTW, how is transaction handled?
SmokeMachine xinming: transactions are NYI 19:31
xinming Ok, But for now, We can still wrap txn ourselves by manually get the dbh 19:32
SmokeMachine xinming: I'm happy you are liking it... :) 19:33
xinming: what parts you are geting more troubles? 19:34
xinming SmokeMachine: Not much for now, maybe in the future, We may try customized sql function call, customized join 19:56
If customized join doesn't work, I can use view to do that, which is not a big problem ATM.
Also customized baseclass for ResultSeq 19:59
Maybe different class/role applied to the Row object based on different logic. 20:00
xinming m: my %h; my %y = %(x => %("a" .. "d")); %h.append(%y<x><a c>:p[], :x); %h.perl.say; 20:35
camelia {:a("b"), :c("d")}
xinming How can I also append the :x to the %h?
If I don't want to do it twice
don't want to append %h twice 20:36
m: my %h; my %y = %(x => %("a" .. "d")); %h.append(%(%y<x><a c>:p, :x)); %h.perl.say;
camelia {"a\tb c\td" => :x}
xinming m: my %h; my %y = %(x => %("a" .. "d")); %h.append(%(|(%y<x><a c>:p), :x)); %h.perl.say; 20:37
camelia {:a("b"), :c("d"), :x}
xinming This one worked, But not perfect.
Any ideas?
I don't like the |(), As it might run out of memory IIRC 20:39
timotimo m: my %h; my %y = %(x => %("a" .. "d")); %h.push(%(|(%y<x><a c>:p), |(:x))); say %h.perl 20:42
camelia {:a("b"), :c("d"), :x}
timotimo it's not shorter 20:43
SmokeMachine m: my %h; my %y = %(x => %("a" .. "d")); %h = |%y<x>, :x; %h.perl.say 20:44
camelia {:a("b"), :c("d"), :x}
xinming timotimo: there is | 20:47
timotimo yes, true
xinming The reason I don't like | is, when you use it with push, It sometimes make program quit directly.
Before I meet once, IIRC, the array is too large.
so, I learnt @array[] and %xx<> 20:48
timotimo sounds like accidentally trying to expand an infinite sequence
or maybe argument lists growing beyond the limit?
xinming timotimo: yes, arg lists growing beyond the limit 20:50
timotimo ah, yes
vrurg m: sub DYNAMIC($name) { say "DYN $name"; 42 }; say $*no-var 20:58
camelia DYN $*no-var
42
vrurg heh... ;)
A solution for the recent weekly. 20:59
SmokeMachine Creating a new file I had a good surprise! :) usercontent.irccloud-cdn.com/file/.../image.png 21:02
I didn't know this DYNAMIC sub... 21:03
vrurg SmokeMachine: this is how $* gets resolved.
SmokeMachine can you belive I got confused by model and module?!! :P 21:06
m: sub DYNAMIC($name) { say "DYN $name"; 42 }; say $*no-var; my $*yes-var = 42; say $*yes-var
camelia DYN $*no-var
Cannot modify an immutable Int (42)
in block <unit> at <tmp> line 1

42
DYN $*yes-var
SmokeMachine m: sub DYNAMIC($name) { say "DYN $name"; 42 }; say $*no-var; my $*yes-var = 13; say $*yes-var
camelia DYN $*no-var
Cannot modify an immutable Int (42)
in block <unit> at <tmp> line 1

42
DYN $*yes-var
SmokeMachine m: sub DYNAMIC($name) is rw { say "DYN $name"; $ //= 42 }; say $*no-var; my $*yes-var = 13; say $*yes-var 21:07
camelia DYN $*no-var
42
DYN $*yes-var
DYN $*yes-var
13
vrurg SmokeMachine: oh, easily. Yesterday I was double-fixing a single simple condition. Messing up two similar words? Easily... ;)
m: sub DYNAMIC($name) is raw { $ //= 42 }; $*no-var = 1; say $*yes-var 21:08
camelia 1
vrurg m: sub DYNAMIC($name) is raw { state %h; %h{$name} }; $*no-var = 1; say $*yes-var; 21:09
camelia (Any)
SmokeMachine vrurg: how is the original implementation? Does it leak memory? 21:24
vrurg SmokeMachine: Sure, no. It only lookups for the requested symbols 21:25
SmokeMachine m: sub DYNAMIC($name) is raw { state %h; %h{$name} }; for ^10 { $*no-var .= 1 }; say $*no-var 21:26
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing required term after infix
at <tmp>:1
------> 3e %h; %h{$name} }; for ^10 { $*no-var .=7⏏5 1 }; say $*no-var
expecting any of:
dotty method or postfix
SmokeMachine m: sub DYNAMIC($name) is raw { state %h; %h{$name} }; for ^10 { $*no-var ~= 1 }; say $*no-var 21:27
camelia 1111111111
SmokeMachine m: {my $*no-var; for ^10 { $*no-var ~= 1 } }; # is this still storing 1111111111? 21:28
camelia ( no output )
SmokeMachine m: sub DYNAMIC(|c) is raw { say c; 1 }; { say $*no-var }; say $*no-var 21:30
camelia \("\$*no-var")
1
\("\$*no-var")
1
vrurg SmokeMachine: don't actually rely on this. I'm considering changing this and getting rid of the sub, moving the functionality into PseudoStash. 21:31
I don't feel good about possible accidental redefinition of the sub in a user code. 21:32
Geth_ ¦ problem-solving: vrurg assigned to jnthn Issue Dynamic symbols resolution via &DYNAMIC is potentially a problem in user code github.com/perl6/problem-solving/issues/127 21:40
rba raku.org is serving github.com/perl6/perl6.org independently NOW. let me know if you find any issues. 21:46
uzl[m] m: class A { has $!n = 1; method n { $!n } }; A.new.n.say 23:03
camelia 1
tellable6 2019-10-15T16:03:09Z #raku <jmerelo> uzl changes are agreed, but not yet deployed. Documentation should reflect what's already there.
uzl[m] .tell jmerelo Thanks for letting me know ;-)! 23:04
tellable6 uzl[m], I'll pass your message to jmerelo
uzl[m] m: class A { has $.n = 1; method new { self.bless() } }; A.new.n.say
camelia 1
uzl[m] So which one of these shows more intent? The end goal is that I don't want the client to set the attribute during construction. The first one declares a $! attribute and creates a method of the same name. The second one declares a $. attribute and then overrides the new method so as to avoid the client setting it. 23:09
jnthn Well, the first isn't really possible to circumvent, whereas the second is by the client calling .bless or even .Mu::new(:1n) :) 23:10
uzl[m] I'm inclined to go with the former, at the expense of creating a method that's already provided by Raku, since I don't have to mess with the object construction needlessly.
jnthn The other option you miss is declaring a `submethod BUILD() { $!n = 1 }`, since BUILD gives complete control of the construction at that inheritance level. 23:11
uzl[m] Yeah, totally forgot about BUILD... ;-) 23:14
uzl[m] m: class A { has $.n; submethod BUILD() { $!n = 1} }; A.new(n => 5).n.say 23:15
camelia 1
uzl[m] Oh, great. It seems I wrongly assumed that the BUILD's signature needed to match whatever the bless method's capture is ( (:$n) in this case). 23:20
Geth_ perl6.org: f37711cb14 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 7 files
Initial Raku changes

Now that raku.org no longer redirects, I think it's time to start changing references to Perl 6. This is the first pass and I manually corrected just some of the pages. It is a bit hard because there are many different variations (with space, with no space, with unicode nbsp, with html entity), and sometimes it is part of “Rakudo Perl 6”, ... (5 more lines)		 23:47
AlexDaniel rba: ↑
AlexDaniel rba: https issue here: docs.raku.org/ 23:51
wrong certificate
rba: same as modules.raku.org/ and everything else

Powered by Cro and the Raku® Programming Language.

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

Thank you!