»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
labster m: rx/<$y=$x>/ 00:21
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in metachar:sym<assert>; couldn't find final '>'
at <tmp>:1
------> 3rx/<$y7⏏5=$x>/
lookatme morning. o/ 00:24
BenGoldberg Evening :)
lookatme :) 00:25
haxmeister I need a simple module for parsing HTML.. any suggestions? 01:24
MasterDuke_ inline::python and beautifulsoup? 01:35
haxmeister gross 01:36
how could you say that word here..lol.. j/k
I want pure perl6
looking at WWW
lookatme haxmeister, Have you see HTTP::Paser already ? 01:41
sorry.. 01:42
HTML::Parser
haxmeister that yours?
lookatme No 01:43
haxmeister github.com/tokuhirom/p6-HTTP-Parser
haxmeister have you used it? 01:45
lookatme No, I mean HTML::Parser. But it's just a role, not implementation :( 01:47
haxmeister ah man 01:48
haxmeister I wonder if an XML parser would pick up html 01:48
lookatme Why they named a role HTTP::Parser... 01:49
BenGoldberg Only if that html happens to be xhtml
haxmeister would be nice to get html into a list type structure like json-tiny does so nicely
haxmeister guess I could just do it the long way.. regex is ridiculously fun anyway..lol 01:51
BenGoldberg Why does it have to be pure perl6?
What's wrong with a library binding, like Gumbo?
lookatme Unfortunately, it's not have a HTML parser module yet. 01:52
haxmeister nothing wrong with it.. just preference really.. easier to read/understand documentation and code
BenGoldberg For that matter, what's wrong with writing: use Inline::Perl5; use Mojo::DOM :from<Perl5>; 01:53
haxmeister I know... but I don't study perl5 or python 01:54
lookatme Yeah, just like me 01:54
I don't know many about perl5
as well as python. 01:55
lookatme haxmeister, have you thought write a that module yourself 01:56
haxmeister I went through some tuts on python and bought a book.. but I need my brackets.. plus I like functional style and nesting functions.. none of which is great in python
python is the opposite of me.. I like perl because I can address the problem the way I understand it in my head 01:57
lookatme New parser with Perl 6 grammar, wonderful
haxmeister even if it's les elegant than the next guy
I need to learn grammars.. 01:58
still learning
wonder about this: github.com/zoffixznet/perl6-WWW
samcv why didn't we bump the subversion for nqp and moarvm again... 01:59
lookatme It's just a HTTPs client. 02:00
samcv cause 2017.04.3 requires a newer version of moarvm and nqp, and moarvm and nqp only have tags up to 2017.04.3
BenGoldberg It parses json too. Just not html ;)
haxmeister reading about grammars.. maybe I'll make one for HTML, or try.. it could be my first attempt 02:07
BenGoldberg Reading the w3 spec for how to properly parse html5 makes my brain hurt. 02:13
haxmeister there could be some value in just pulling certain elements from an HTML document too 02:20
like making HTML::tables... just pulls the tables out of any html document makes a nice data structure out of them
haxmeister I know of many many websites that have large amounts of valuable data in tables 02:22
lookatme Em. 02:23
haxmeister could even make it only pull tables with a given id="whatever" or any attribute specifically
that actually sounds fun
lookatme yeah, you can 02:25
haxmeister or divs.. or divs with a specific attribute.. 02:27
that all might be simpler for a noob like me.. but this grammar tut is hard to get my head around 02:28
haxmeister docs.perl6.org/language/grammar_tutorial doesn't give an example data file to show what the text we are parsing looks like.. makes it tough to follow 02:30
lookatme You can ref some module source code 02:32
haxmeister it's gotta be really dumb... like entry level grammar stuffs
new concepts are hard
hashed filenames?... Larry Wall your fired...lol j/k 02:36
lookatme Write some sample, help to understand grammar concept 02:43
perlawhirl is there a way to declare that all arguments to a sub should be the same type. 05:12
so I don't have to do things like: sub foo( Long::Class::Name $x, Long::Class::Name $y, Long::Class::Name $z ) { ... }
Slightly less typing could be: sub foo( $x, $y, $z where all($x, $y, $z) ~~ Long::Class::Name ) { ... }
but was hoping for something along the lines of: sub foo( Long::Class::Name( $x, $y, $z ) ) { ... } 05:13
japhdotcom anyone have a little document/tutorial on debugging nqp? 07:12
J0001 hi 07:17
:)
japhdotcom for example, in Perl6/Grammar token statement_control:sym<use> there is <version>... Can I access <version> from Perl6::Actions method statement_control:sym<use>($/)? 07:24
moritz japhdotcom: that's in $<version> 07:27
moritz in the action method 07:27
japhdotcom moritz, in the code i'm testing... i have use Terminal::ANSIColor:ver('0.2')... yet in the method statement_control:sym<use>($/) { 07:38
nqp::say( $<version> );
is returning blank...?
moritz japhdotcom: check with --target=parse if it actually parses the :ver('0.2') as <version> 07:46
moritz because to me, it looks like the grammar path you're looking at is the one that parses "use v6;" 07:47
and the :ver('0.2') might be part of the module name 07:48
daxim <github.com/ugexe/zef/issues/161> what's the programming language (runtime) got to do with it? to me, this is clearly an application level issue. what do yall think? 07:49
moritz zef presumably uses MAIN() subs for option parsing, so it's a rakudo/design issue indeed 07:53
(and we don't have a glob() yet either)
lookatme haxmeister, I made a simple grammar for html format gist.github.com/araraloren/379f1c8...2ccf602c70 08:12
You can modify the html file, see how it works. 08:13
AlexDaniel huggable: parse HTML 08:38
huggable AlexDaniel, nothing found
AlexDaniel huggable: parse HTML :is: Use Gumbo. github.com/Skarsnik/perl6-gumbo
huggable AlexDaniel, Added parse HTML as Use Gumbo. github.com/Skarsnik/perl6-gumbo
japhdc moritz, you are correct... when checking --target=parse, I was comparing it to the wrong grammar path... looks like it gets parsed as something like... identifier "ver" EXPR value 0.2 is 09:07
japhdc moritz, from inside method statement_control:sym<use>($/) the farthest down available parse item is $<module_name> which gets further parsed... from inside the :sym<use>($/) method is there a way to access those further down components? 09:10
moritz japhdc: I'd hope that $<module_name>.ast has the module name dissected into a usable object somehow 09:42
japhdc: you should avoid digging deep down in nested parse trees in action methods, because that causes tight coupling of the action methods to the grammar structure 09:43
japhdc moritz: i'm trying to implement a very friendly autocompletion function... using the action methods as starting points for building a list of what is available seems like the most direct path... unless you have other ideas? 09:45
moritz japhdc: no, seems fine 09:46
japhdc: I'm just warning against digging too deep down into the parse tree from action methods 09:47
avoid more than one level where possible/practical
japhdc moritz: could you give an example or two of what I can get out of $<module_name>.ast ? 09:49
unfortunately it doesn't seem that "dd" will work inside of nqp... so once you have an object that you aren't familiar with, any shortcuts to inspecting it? 09:51
moritz you can find out what type it is with $obj.HOW.name($obj) 09:52
but I'm not aware of a general dumper in p6 land 09:53
though others (possibly in #perl6-dev) might have better tricks
Geth doc: 2af18ad4d3 | (Zoffix Znet)++ | CONTRIBUTING.md
Remove Inline::Python stuff

Fixes #1301		 10:17
tadzik m: say "foo" ~~ s/foo/bar/r; # how do I do this in P6 again? 10:28
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3say "foo" ~~ s/foo/bar/7⏏5r; # how do I do this in P6 again?
expecting any of:
infix
infix stopper
postfix
sta…
tadzik oh, subst. 10:29
lizmat yup 10:33
tadzik oh dear. Segmentation fault, in MVM_profile_instrumented_mark_data 10:37
nine .tell skids with github.com/skids/perl6sum/commit/c...7be2702906 you removed the META.info, but the META6.json does not have a source-url, so the module is no longer installable 10:41
yoleaux nine: I'll pass your message to skids.
tadzik samcv++ Zoffix++ # tons of good PRs for my METAfiles :) 10:50
Zoffix nine: yeah, it's there in support -> source 10:54
And it fetches just fine 10:56
.tell skids the META is fine, but there are test failures due to "P6M Merging GLOBAL symbols failed: duplicate definition of symbol $synopsis" 10:57
yoleaux Zoffix: I'll pass your message to skids.
nine Zoffix: so is support/url the official fallback for source-url? 10:58
Zoffix: why do we have 2 fields for essentially the same?
Zoffix nine: support -> source is the "official" field and "source-url" is the ecosystem extension. Why have two? I guess same reason we had META.info vs META6.json; someone used the pre-Christmas format and someone else just copied it over 11:00
Geth ecosystem: ae420f0163 | (Tadeusz Sośnierz)++ (committed using GitHub Web editor) | META.list
Remove Task::Star

It was constantly outdated, and more confusing than helpful apparently.		 11:11
cygx I (incorrectly?) assumed source-url is for programs and support/source is for humans... 11:15
first time I've heard or support/url 11:16
tadzik source-url evolved from repo-url, which was the name I made up back in neutro days, I think 11:17
then support/source I think was adopted from the module synopsis?
pmurias where is our META6.json spec? 11:20
tadzik design.perl6.org/S22.html#META6.json it's here 11:21
it says that it may be out of date and to read doc.perl6.org instead, but doc.perl6.org links to this :P
pmurias it has an outdated warning
Zoffix pmurias: all pages have that warning 11:22
tadzik "Warning: This warning may be out of date"
Zoffix The only thing outdated about S22 is the resources format IIRC
ugexe s22 still references CompUnitRepository
not CompUnit::Repository 11:23
Zoffix None of the META stuff is in roast to the best of my knowledge
Nad there's an open doc Issue about META; about poor information on what is "spec" and proper 11:24
Zoffix nominates ugexe to fix all that :)
pmurias the modules guide mentions source-url and the S22 doesn't
ugexe because source-url is an ecosystem spec field
Zoffix That Issue: github.com/perl6/doc/issues/764 11:25
pmurias what's an ecosystem spec field? 11:27
tadzik hm, github.com/perl6/ecosystem/blob/ma...atelist.pl is still Perl 5 11:29
what an opportunity :)
ugexe pmurias: consider putting a module on cpan. you probably want source-url to be whatever you said it was, and let pause/cpan fill in the source-url with wherever it provides it 11:37
ugexe er, you probably want "support" : { "source" : " ..." } to be whatever you declared rather 11:38
tadzik I'm thinking about building something for modules.perl6.org that looks at modules.perl6.org/update.log and has a place for the module authors to easily look up deficiencies in their modules. Does something like that exist already? 11:39
Geth Swapped META.info → META6.json in 13 dists in github.com/perl6/ecosystem/commit/a709bf360f 11:42
tadzik nine: this should fix Acme::Meow :) 11:44
so ecosystem has update.pl, updatelist.pl and server/updatelist.pl 11:54
and modules.perl6.org has its own update-modules.perl6.org, with a separate codebase for that
so are the ecosystem scripts even used anywhere at this point?
(any of the three? :))
moritz yes 12:01
server/updatelist.pl is used for updating ecosystem-api.p6c.org/ 12:02
otoh I have no idea if anybody still uses ecosystem-api.p6c.org/ 12:03
because panda seems to be mostly dormant, and iirc zef has its own thing
tadzik aha
nodnod
moritz winkwink 12:05
pmurias so source-url would point not to the repo but to the place you can fetch the module? 12:07
tadzik that's the idea, I think 12:08
cygx perhaps also of interest: github.com/jonathanstowe/META6/blo...L225..L246 12:10
pmurias it would be great to have everything specced down, you can't expect anybody to conform to stuff that's not written down 12:11
nine If only there were a sort of conference where the relevant people could sit down and get this nailed... 12:19
timotimo :D
El_Che nine: you're famous 12:22
I saw your pic on twitter
It looked like you were presiding over a religious service
did you?
pbs.twimg.com/media/C_namDvXcAAyyWV.jpg:large 12:23
nine El_Che: haha :) 12:28
pmurias nine: any Perl 6ers coming to The Perl Conference in Amsterdam? ;) 12:33
eythian oh, I need to see about going to that 12:35
El_Che it would be great to have some pricing info, so I can start to convince work to send me :) 12:37
nine pmurias: I definitly am going to 12:38
tadzik I think I'm going too
nine \o/
eythian I live in Amsterdam, so definitely should.
El_Che eythian: you have zero excuse now :) 12:39
eythian sha2017.org/ this finishes the day prior, it'll be quite a week all up :)
El_Che eythian: The last few years I seem to combine FOSDEM and cfgmgmtcamp. Pretty tired after those 4 days 12:44
eythian I also have a beer festival to go to the day after the perl conf, so that'll be my relaxation... 12:44
nine Found another one! Text::Tabs uses source-uri instead of source-url in it's META6.json 12:47
sena_kun nine, I'll fix it now, thanks. 12:48
nine fired of PRs for all remaining unfetchable dists 12:53
tadzik ugexe: gist.github.com/tadzik/50c0ee877f9...1aba168b86 that's the same thing that happened to me yesterday. Fresh rakudo, fresh zef 12:56
tadzik it doesn't seem like it realized MIME::Base64 is LWP's dependency 12:56
and then 'zef install LWP::Simple' says 'All candidates are currently installed' 12:57
pmurias has hotel/flight to TPC booked, I'll very likely submit a rakudo.js talk
tadzik yay! 12:58
stmuk pours one out for Task::Star
tadzik plays the sad trumpet 12:59
timotimo m: use Test; sub fails { fail "oh no" }; isa-ok fails, Failure, "fails fails"; use nqp; nqp::force_gc(); nqp::force_gc(); 13:00
camelia ok 1 - fails fails
timotimo oh
m: use Test; sub fails { fail "oh no" }; isa-ok fails, Failure, "fails fails"; use nqp; nqp::force_gc(); nqp::force_gc(); nqp::force_gc();
camelia ok 1 - fails fails
tadzik I'm thinking about making Task::Aster that'll just contain a subjective list of cool stuff
timotimo ...
m: use Test; sub fails { fail "oh no" }; { isa-ok fails, Failure, "fails fails"; }; use nqp; nqp::force_gc(); nqp::force_gc(); nqp::force_gc();
camelia ok 1 - fails fails
timotimo :\
Geth ecosystem: 2c20a921f3 | (Zoffix Znet)++ (committed using GitHub Web editor) | META.list
Remove Ecosystem-Test
 13:01
stmuk Zoffix's camelion has a cool logo :) 13:02
"logoware" :)
timotimo anyway, what i was trying to demonstrate was that isa-ok won't disarm failures
stmuk if he doesn't get sued by suse 13:03
timotimo should we a) make isa-ok check .defined on objects that get passed in, b) make method isa on a failure disarm it if checked against Failure explicitly, c) something else?
pmurias what's camelion? 13:07
timotimo something regarding tagging of modules 13:08
see perl6.party for a post about it
[Coke] struggles through his 2nd coffee. 13:15
smls m: my %hash = "a" => 1; my (:$a!, :$non-existent-key!) := %hash; 13:18
camelia MVMArray: atpos expected string register
in block <unit> at <tmp> line 1
smls bisectable6: my %hash; try my (:$non-existent-key!) := %hash; say "$!";
bisectable6 smls, Bisecting by output (old=2015.12 new=c240195) because on both starting points the exit code is 0
smls, bisect log: gist.github.com/e294eb15a944ceaab5...5cf69e6598
smls, (2017-02-12) github.com/rakudo/rakudo/commit/b2...8dad8c0632
smls ^^ regression in error handling for required named parameters 13:22
lizmat smls: testing a fix now
lizmat smls: please pull - github.com/rakudo/rakudo/commit/95c5e1ffa8 13:26
raschipi Zoffix: Give us a link to Camelion 13:32
tadzik Zoffix: ping
Zoffix Backlogging a bit, and seems people aren't aware we *do* have a nice and shiny, pure-perl HTML parse: DOM::Tiny modules.perl6.org/dist/DOM::Tiny
tadzik I need help with JSON::Meth and what it does in mpo code 13:33
Zoffix Author says it's alpha-quality, but it's a port of a polished Perl 5 module, so I'd say API is safe to use and it worked fine when I last used it
tadzik: exports $j variable
.tell haxmeister I found this HTML parsing module work just fine: modules.perl6.org/dist/DOM::Tiny And WWW is just a user agent, not a parser 13:34
yoleaux Zoffix: I'll pass your message to haxmeister.
tadzik Zoffix: in lib/ModulesPerl6/DbBuilder/Dist/Source.pm it does $data->$json, that, I guess, parses $data as json and makes this $json var globally visible?
Zoffix 1 sec
raschipi: github.com/zoffixznet/camelion/blo...s/logo.png 13:35
smls lizmat: Your commit seems to do the trick, thanks!
Zoffix tadzik: ok, in that code, it exports $json because $j is too non-obvious. It decodes $data to JSON if it's JSON and encodes it into JSON if it's not JSON 13:36
tadzik: is this the code you're asking? github.com/perl6/modules.perl6.org...rce.pm#L74
tadzik yes
so now the global $json variable contains decoded $data?
Zoffix tadzik: it's equivalent to Perl 6's; `use JSON::Fast; try from-json $data; if $! { log error => "Failed to parse: JSON error: $!"; return; } 13:37
tadzik: no, it's just a method name
tadzik: metacpan.org/pod/JSON::Meth
tadzik but in #80 you write to it
yeah, I read the pod, it didn't help :P
Zoffix tadzik: oh wait, yeah it does
tadzik: OK, now I remember. Yeah, $json will then contain decoded json 13:38
Zoffix tadzik: it's equivalent to Perl 6's; `use JSON::Fast; try my $json = from-json $data; if $! { log error => "Failed to parse: JSON error: $!"; return; }; $json{do things} 13:38
tadzik: why? Is there a problem you're trying to fix? 13:39
tadzik Zoffix: I try to make it so that the warnings/errors in modules.perl6.org/update.log are available as something more structured as a logfile
so that I can go to mpo or somewhere and it'll tell me: these modules of yours don't have tags, or so 13:40
Zoffix tadzik: oh, that's very easy to do
tadzik: all the messages in update.log are emitted using github.com/perl6/modules.perl6.org...der/Log.pm
Zoffix tadzik: so you could make it also gen a JSON file or something 13:41
tadzik yeah, but this log doesn't know for what module this is reported
Zoffix Ah, right
tadzik I made it so that the dists contain problems just like they now contain travis_status
so we can store that and look it up later
Zoffix tadzik: I see there's already a "META checker post-processor": github.com/perl6/modules.perl6.org...Checker.pm 13:43
tadzik yep
Zoffix Sounds like a prime place to add the tag checking or whatever
tadzik but most of the warnings come from Source.pm
perhaps METAChecker would be a better place to put them instead 13:44
Zoffix Maybe
.tell lookatme we do actually have an HTML parser; DOM::Tiny: modules.perl6.org/dist/DOM::Tiny 13:45
yoleaux Zoffix: I'll pass your message to lookatme.
Zoffix tadzik: if you need more help, just ping me in #perl6-dev
\o
tadzik will do 13:45
nine So many ecosystem modules are missing license information in their META6.json :/ 13:54
[Coke] nine: samcv was working on that, no? 13:58
tadzik yep, I got 15 PRs for it I think 14:00
raschipi It was way worse before samcv had a go at it. 14:03
She took it from ~5% to more than 50%.
ugexe tadzik: i commented on your gist... I can't reproduce. In my case it finds everything fine, but freezes up testing GTK::Simple 15:38
ugexe and if an install fails due to precomp it remains installed 15:39
Geth ecosystem: 76c34d269a | (Tony O'Dell)++ (committed using GitHub Web editor) | META.list
Adding System::Query to META

A utility for collapsing JSON dependent upon certain environmental information (useful for BUILDing/dependency requirements/probably other things).		 16:11
Geth doc: 12b2675a03 | (Jan-Olof Hendig)++ | doc/Type/Blob.pod6
Added docs for Blob.reverse
 17:18
TreyHarris m: 3 ± 1 18:01
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 0337⏏5 ± 1
expecting any of:
infix
infix stopper
statement end
statement modifier
statement modifier loop
TreyHarris weird, I thought arbitrary precision was implemented
[Coke] That operator doesn't appear in rakudo/src or shown as impl. in docs. 18:05
TreyHarris It's not even in specs.... or in the design archive. Did I totally imagine it? Was it in Pugs?
[Coke]: yeah, that's what I meant
TreyHarris I have old code from 2007 that has ±, so I think it must've been in Pugs 18:05
perlpilot
.oO( Why isn't ± there yet though? ;-) )
18:06
TreyHarris is there a good ack-like searcher than can squash paragraphs together? I hate it when I discover I missed something I searched for in docs because the pattern spanned a line break 18:10
perlpilot one would think that ack could do paragraph or slurp mode since it's relatively easy to do in perl. 18:13
Geth doc: 2131bc2cc5 | (Jan-Olof Hendig)++ | doc/Type/CallFrame.pod6
Fixed broken example. Zoffix++, PerlPilot++
 18:15
TreyHarris perlpilot: closed issues say ack is too-deeply line-oriented to implement 18:18
hobbs TreyHarris: ag docs say "--[no]multiline: match regexes across newlines. Enabled by default." 18:20
so that seems promising
perlpilot TreyHarris: maybe try the silversearch or ripgrep to see if they do it
hobbs++
perlpilot has taken to use ripgrep lately 18:21
hobbs I don't know that one 18:21
TreyHarris hobbs++ 18:22
perlpilot hobbs: github.com/BurntSushi/ripgrep
hobbs yeah, looking at it. 18:23
TreyHarris I just tried agrep (which allows arbitrary-delimeters but is single-file/stream-oriented like *grep, but if you give it a delimiter (in this case '$$'), you can't use regexes, only literal strings
But ag works, thanks.
hobbs rg specifically lists not having multiline searches in its caveats 18:24
hobbs otherwise looks nice though :) 18:24
perlpilot yeah ... I'm not even sure anymore why I use ripgrep over ag (but both are blazingly fast on my large and deeply nested directories, which is why they get favored over ack for many things) 18:25
perigrin pets ag 18:26
TreyHarris ah, before I saw it doesn't support multiline I tried to install it (both via repo and from local source, using the CLI's given in the README). Didn't work, but I've never used Rust before so there may be something blindingly obvious wrong here: gist.github.com/treyharris/511117b...4d00a103f7 18:30
TODO[@Learning|@Coding]: Rust exercism 18:31
gah
ajr_ Is there a problem with the mail server at perl6.org? I tried to file a rakudobug this morning: [email@hidden.address] bounced. 18:50
raschipi ajr_: If you get a gist or pastie of the bounce with your personal data removed, I will have a look at it. 18:52
ajr_ The response from the remote server was: 554 5.7.1 [email@hidden.address] Recipient address rejected: Access denied 18:54
cygx ajr_: isn't it [email@hidden.address] 18:57
TreyHarris cygx: yes
ajr_: ^^
curt_ m: say 1 ∈ (Set) ?? 'present' !! 'not present'; 18:59
yoleaux 15:15Z <Zoffix> curt_: Our modules.perl6.org currently rejects logotypes that aren't 32x32px. Sent you a PR to fix yours: github.com/CurtTilmes/perl6-eredis/pull/1
camelia MoarVM panic: Memory allocation failed; could not allocate 83968 bytes
ajr_ cygx: Oops, yes. Thank you. 19:02
ajr_ Does anyone know how many people are experimenting with Perl 6 on the Pi? 19:06
MasterDuke_ ajr_: i know several have, but i couldn't tell you all the names. maybe try searching the irc logs? irclog.perlgeek.de/perl6/search 19:08
Zoffix curt_: it infini-loops in dispatch because Set.Set returns Set. And so far I see (elem), (cont), ∉, ∋, ∌ similarly affected. There might be more. Would you mind filing a bug for it? 19:16
huggable: rakudobug
huggable Zoffix, Report bugs by emailing to [email@hidden.address]
Zoffix And some ops like ⊇ and ⊉ crash with "can't look up attributes... " error. There's probably a meaning for type objects we could implement; like Set eqv set() 19:20
ajr_ Huggable: Zoffix doesn't report bugs, he fixes them. :-)*
curt_ rt.perl.org/Public/Bug/Display.html?id=131300 19:21
Zoffix TreyHarris: what does 3 ± 1 mean? 19:21
curt_: thanks
ingy does p6 have global vars, or something like them? 19:23
google is not my friend on this so far
ajr_ exit
Zoffix ingy: yeah, PROCESS:<$SOME-VAR>
m: sub make-one { PROCESS:<$SOME-VAR> := 42 }; say $*SOME-VAR // 'not there'; make-one; say $*SOME-VAR 19:24
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix := instead
at <tmp>:1
------> 3sub make-one { PROCESS:<$SOME-VAR> :=7⏏5 42 }; say $*SOME-VAR // 'not there'; ma
Zoffix uhh
Zoffix quickly glances at sauce
perlpilot ingy: I went to doc.perl6.org and typed 'GLOBAL' into the search box and got docs.perl6.org/language/packages#i...package%29 19:24
cygx PORCESS::
Zoffix ah
m: sub make-one { PROCESS::<$SOME-VAR> := 42 }; say $*SOME-VAR // 'not there'; make-one; say $*SOME-VAR
camelia not there
42
Zoffix sweet
perlpilot ingy: depending on what you mean by "global", one of those is sure to be useful to you
Zoffix ingy: basically PROCESS:: is the last place lookup of dynamic vars is done. So if you stick the value there, in any other place that dynamic will now have a value, unless it's overriden somewhere up the call stack (e.g. with `my $*SOME-VAR`) 19:25
m: sub infix:<±> { .min .. .max with minmax $^a-$^b, $a+$b }; say 3 ± 1 19:29
camelia 2..4
Zoffix m: sub infix:<±> { .min .. .max with minmax $^a-$^b, $a+$b }; say -3 ± 1
camelia -4..-2
Zoffix m: sub infix:<±> { .min .. .max with minmax $^a-$^b, $a+$b }; say (3 ± 1) ~~ 3.5
camelia False
Zoffix m: sub infix:<±> { .min .. .max with minmax $^a-$^b, $a+$b }; say 3.5 ~~ (3 ± 1)
camelia True
Zoffix hehe
ingy thanks all 19:32
TreyHarris Zoffix: good question, turns out when I just went down the rabbit hole and tried to see what other languages do, that it depends on context. Some languages would take it to mean 2..4; some would take it to mean a standard deviation of mean 3 and sigma 1, a.k.a. an "error form"; and some take it to mean a arbitrary-fixed precision number such that 3±1 is 3 with an epsilon of 1/3 (intuitively I think that 19:41
means what I think of as an epsilon of 1, but apparently in fixed-precision arithmetic you define epsilon as a coefficient of the value, not a variance of the value)
(the latter means that x±y when y ≥ x is NaN) 19:44
perlpilot TreyHarris: mayhap you were thinking of the ≅ operator with a tolerance (before, when you were expecting ± to work) ? 19:46
TreyHarris perlpilot: pretty sure i was. 19:47
atweiden i'm trying to group an array of DateTimes with overlapping Dates, by Date 21:29
any inspiration?
timotimo so, like, using categorize? 21:32
atweiden interesting 21:35
i may try that
Geth doc: e1281f2816 | (Jan-Olof Hendig)++ | doc/Type/Iterable.pod6
Fix typo
 21:40
timotimo there's another one just like categorize
classify
perlpilot yeah, but categorize is the right one in this case I think. 21:41
timotimo Note that, unlike classify, which assumes that the return value of the mapper is a single value, categorize always assumes that the return value of the mapper is a list of categories that are appropriate to the current value.
i think you want classify
perlpilot huh, you and I got different things from what he wanted
The "overlapping dates" business makes me think categorize is right. 21:42
timotimo every DateTime only has one Date, though? 21:43
atweiden it's like [2016-01-01T12:35Z, 2016-01-01T15:55Z, 2017-12-31T01:01Z]
combine 2016-01-01's
it's not that simple, as the datetimes are part of a class, and i will need to combine the classes 21:44
perlpilot ah, I guess I was thinking of date ranges rather than just dates 21:45
timotimo++

Powered by Cro and the Raku® Programming Language.

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

Thank you!