🦋 Welcome to the former MAIN() IRC channel of the Raku Programming Language (raku.org). This channel has moved to Libera (irc.libera.chat #raku)
Set by lizmat on 23 May 2021.
00:03 reportable6 left 00:06 reportable6 joined 00:29 Kaipi joined, Kaipi left 00:43 swaggboi joined 01:02 RandalSchwartz joined 01:55 monkey_ joined 02:03 monkey_ left, Manifest0 left 02:06 Manifest0 joined 02:50 monkey_ joined 03:42 swaggboi left 03:45 swaggboi joined 03:56 monkey_ left 04:56 squashable6 left, benchable6 left, evalable6 left, statisfiable6 left, greppable6 left, linkable6 left, notable6 left, nativecallable6 left, tellable6 left, quotable6 left, sourceable6 left, coverable6 left, unicodable6 left, releasable6 left, bloatable6 left, bisectable6 left, shareable6 left, committable6 left, reportable6 left, nativecallable6 joined, committable6 joined, tellable6 joined 04:57 coverable6 joined, shareable6 joined, notable6 joined, squashable6 joined, unicodable6 joined, sourceable6 joined 04:58 quotable6 joined, benchable6 joined, releasable6 joined 04:59 statisfiable6 joined, evalable6 joined, bisectable6 joined, bloatable6 joined, greppable6 joined, linkable6 joined, reportable6 joined 05:33 squashable6 left 05:34 squashable6 joined 06:02 reportable6 left 06:04 reportable6 joined 06:20 frost joined 06:32 RandalSchwartz left 06:38 guifa left 07:03 frost left 07:09 frost joined 08:09 coverable6 left, unicodable6 left, sourceable6 left, benchable6 left, squashable6 left, tellable6 left, greppable6 left, bisectable6 left, quotable6 left, linkable6 left, nativecallable6 left, notable6 left, shareable6 left, statisfiable6 left, evalable6 left, committable6 left, releasable6 left, bloatable6 left, reportable6 left, releasable6 joined, statisfiable6 joined, sourceable6 joined 08:10 unicodable6 joined, notable6 joined, evalable6 joined, shareable6 joined, committable6 joined 08:11 quotable6 joined, bisectable6 joined, reportable6 joined, nativecallable6 joined, linkable6 joined, greppable6 joined, squashable6 joined, tellable6 joined, benchable6 joined 08:12 bloatable6 joined, coverable6 joined 09:04 MasterDuke34 left 09:11 MasterDuke joined 09:19 MasterDuke left 09:29 MasterDuke joined 09:33 Sgeo left 10:05 stoned75 joined 10:09 Manifest0 left 10:13 Manifest0 joined 10:26 holly_ joined 10:33 stoned75 left 10:52 stoned75 joined 11:10 stoned75 left 11:13 domm left 12:03 reportable6 left 12:05 reportable6 joined 12:14 yewscion joined 12:46 leah2 left 12:47 leah2 joined 12:52 stoned75 joined, stoned75 left 13:34 b2gills left 13:44 domm joined 14:00 antononcube joined
antononcube Hi! I want to define left and right "word" boundary assertions that for which "word" consist of alpha-numeric characters _and_ punctuation characters. E.g. I want to have word boundaries for words like ".net" or "c++" and parse a list of such words. I want to parse a list of 14:08
This left boundary definition seems to work together with >> : regex wbpl { <!after [ <alnum> | <punct> ]> <?before [ <alnum> | <punct> ]> }
The parsing stops working when I try to use this right boundary defintion: regex wbpr { <?after [ <alnum> | <punct> ]> <!before [ <alnum> | <punct> ]> }
Any suggestions?
moritz antononcube: what's an example of where you want it to match and where it doesn't match? 14:09
antononcube '.net framework, c++, java, and software architect' 14:10
@moritz Thank you for replying! Please, take a look in these definitions: github.com/antononcube/Raku-DSL-En...es.rakumod 14:11
moritz antononcube: perlpunks.de/paste/show/60c61269.6e90.d6 your word boundary regex seems to work for me 14:13
perlpunks.de/paste/show/60c612d3.6e90.3b3 this one illustrates all the position where it matches 14:14
antononcube @moritz Ok, good to know! 14:15
I am trying it out now.
14:26 yewscion left 14:28 holly_ left
antononcube @moritz Thank you! You comment really helped -- I found a mistake in another part of my program... 14:29
I have another question -- it is about using Raku in Babel org mode. (orgmode.org/worg/org-contrib/babel/) 14:34
Does someone here use Babel org-mode and Raku? I read somewhere that Raku is supposed to supported in Babel, but I do not think it is. 14:35
I found this emacs-Lisp definition: mail.gnu.org/archive/html/emacs-de...00636.html . And with some little tweaking I made it work in Aquamacs. It looks like though the author, Tim Van den Langenbergh, has abandoned any efforts in that direction and deleted / removed the corresponding files from hist GitHub repository. 14:39
14:39 holly_ joined
gfldex lolibloggedalittle: gfldex.wordpress.com/2021/06/13/te...-waterbed/ 14:42
14:43 RandalSchwartz joined 14:47 yewscion joined
gfldex moritz: I'm on page 40 of "Parsing with Perl 6" and I already learned a few new tricks. :) 14:48
14:51 yewscion left 14:56 frost left
moritz gfldex: glad you're getting something out of it :D 14:57
antononcube @gfldex Yes, it is a good book -- I learned a lot about Raku grammars from it.
moritz bows and blushes
14:57 b2gills joined
Altreus Perl 6␛b6xA Raku <- this puts Perl Raku 14:59
</pedantry>
Altreus would use FPCRaku 15:00
oh, no it doesn't because you didn't put a space after the 6!
gfldex I tried to shoehown a 6 into that vim command.
Altreus :) 15:01
15:02 domattps joined
antononcube I do not like this vim-direction in the discussion. I am interested in using Raku in Emacs... :/ 15:04
Altreus gfldex: you've put where instead of were a couple of times btw
gfldex Thanks for the denitting. 15:05
15:25 domattps is now known as GoneDomattps 15:27 GoneDomattps is now known as domattps 15:41 antononcube left 15:53 lgtaube left 15:58 domattps is now known as GoneDomattps 16:02 GoneDomattps is now known as domattps 16:09 lgtaube joined 16:12 domattps is now known as GoneDomattps 16:18 Sgeo joined 16:19 tobs left 16:21 tobs joined 16:29 Manifest0 left, Manifest0 joined 16:34 antononcube joined 16:35 antononcube left 16:36 Manifest0 left, Manifest0 joined 16:47 holly_ left 16:48 Manifest0 is now known as testingstuff 16:49 testingstuff is now known as Manifest0
SmokeMachine I've started writing this module to test my Pod code snippets, does anyone have any advice of a better way of doing that? Or would like to help me on evolving that? (Sorry, there is no good README yet...) github.com/FCO/Pod-Test-Code 17:04
17:49 evalable6 left, linkable6 left, evalable6 joined 17:50 linkable6 joined 17:54 abraxxa-home joined 18:03 reportable6 left 18:04 reportable6 joined 18:28 cbk joined 18:56 cbk left 19:01 squashable6 left 19:02 squashable6 joined 19:14 abraxxa-home left 19:27 Xliff joined
Xliff Hello! 19:27
I am now getting this error:
Cannot modify an immutable Block (-> |c { #`(Block|948...)
at /home/cbwood/Projects/p6-GIO/lib/GIO/Roles/AppInfo.pm6 (GIO::Roles::AppInfo):429
That error points here: github.com/Xliff/p6-GIO/blob/maste...o.pm6#L429 19:29
This is a regression from last week's Raku.
This error was not reported in g0de28ae72. 19:30
Can someone take a quick load at that code and see if they can find an eror. Note I have used this code in previous raku, no proble. 19:31
19:35 RandalSchwartz left
Xliff This error appears again, here: 19:43
Cannot modify an immutable Block (-> |c { #`(Block|941...)
at /home/cbwood/Projects/p6-GUPnP/lib/GUPnP/Roles/ACL.pm6 (GUPnP::Roles::ACL):124
lizmat could you golf this to something bisectable, or do you know the offending commit already? 20:05
ah, I should drop the g :-) 20:06
Xliff: can you provide a full stack trace ?
Xliff lizmat: This is over 500,000 lines of code. Where do you golf? 20:07
Am attempting to find the commit, but that is time-consuming and might not finish, today.
Full stack trace incoming
lizmat ah, you're saying that 0de28ae72 is not it 20:08
?
20:08 linkable6 left
Xliff No. Checked that one and it produced the same error. 20:08
It's within the last 3 weeks. Thought last week had a no error week.
I'll have to check the stat files. 20:09
I might need to install bot code to get this one.
lizmat ++Xliff 20:10
20:11 linkable6 joined
raydiak generally you would golf it by making a copy of the project, then rippinng out giant chunks one at a time, checking if the failure still exists after each one. repeat until the only thing left is the failing code, then reduce that to the shortest possible thing you can write which still triggers the bug 20:12
Xliff radiak: Yep. I get the gist of that, but what happens when you are 3 projects into a dependency chain? 20:15
raydiak start by removing the dependencies. if you're lucky, the bug will still exist without them 20:16
Xliff Can't remove the dependencies and expect the code to compile! 20:17
raydiak remove the part that won't compile without them
Xliff Unfortunately that is most of the code
So I have to show examples from the code. Golfing doesn't work with what I have. 20:18
Especially this part. It's more types than anything else.
If you use a type from a previous dependency (which is all over the code.... see GObject)
raydiak remove the type constraints
Xliff You have to include the dependencies
Humm.... I'll think about that.
lizmat Xliff: full stack trace ? 20:19
Xliff But what if the dependencies are a part of the issue?
lizmat: Still working on dependency installs.
raydiak then you at least know that, which is a start
lizmat ok, whenever you have one :-)
raydiak then you start ripping apart the dependencies too 20:20
lizmat raydiak: I think Xliff gets the concept of golfing :-)
raydiak he asked. I answered. 20:21
Xliff radiak: At that point, it's probably just easier for someone to download the code and look at it.
raydiak I looked. didn't see anything obvious. if it's 500,000 lines and the error is elsewhere, that's going to be a tall order 20:22
anyway, apparently I'm not being helpful. going to move on with my morning. good luck, I hope someone can help you 20:23
Xliff Now you feel my pain.
Fortunately at the GIO level it's not over 100k
lizmat: stack trace generated. Creating a gist. 20:27
lizmat: gist.github.com/Xliff/b9067702a014...98ee8f3f79 20:28
lizmat Ah, looks like Method::Also is involved 20:31
does it happen at compile time ? 20:33
Xliff Looks like it. 20:34
And o crap. I think I know why. I made recent changes.
Will the use of "but" modify a block?
If that is true, then I will need to wrap the block in a sub. *sigh* 20:35
lizmat yup :-( 20:36
Xliff Maybe this: 20:37
my \me = -> |c { o."{ $p.value.name }"( |c.list.skip(1), |c.hash ) };
me = me but Method::Also::AliasedMethod;
So could that be wrapped in a sub?
m: role Subby { }; my $a = sub { say "Hi!" } but Subby; $a.^name.say 20:38
camelia Sub+{Subby}
Xliff Cool!
vrurg Xliff: the error rather points at an attempt to assign to a non-Scalar.
Xliff vrurg: Then wouldn't \me count?
vrurg Xliff: sure it will. 20:39
\me becomes a binding to a Block.
You don't need a sub, perhaps. Just use $me =
Xliff Hah! OK. Then changing it to $me should fix it. Checking. 20:40
vrurg As a matter of fact, you can still use \me, but then `me := me but ...` 20:41
Ah, pardon me, `me := ` won't work. :( Though I always expect it to be ok. 20:42
m: my $r := { say "foo" }; role R { }; $r := $r but R 20:43
camelia ( no output )
vrurg Anyway, the above is faster and more memory-efficient.
Xliff vrurg: What about "my $me := -> |c { o."{ $p.value.name }"( |c.list.skip(1), |c.hash ) } but Method::Also::AliasedMethod;" 20:47
vrurg Xliff: Then just `\me = ` will do too, if you prefer this form. 20:48
Xliff I attempted that version of \me and it still generated the error.
vrurg Xliff: I'd say an attempt to assign into `me` is lurking somewhere. 20:49
moritz m: my $r := { say "foo" }; role R { }; $r does R; say $r
camelia -> ;; $_? is raw = OUTER::<$_> { #`(Block+{R}|79142800) ... }
vrurg m: role R { }; my \me = -> |c { } but R; 20:50
camelia ( no output )
vrurg m: role R { }; my \me = -> |c { } but R; say me
camelia -> |c { #`(Block+{R}|87593896) ... }
Xliff cd 20:53
LOL!
raydiak Xliff last note before I'm afk: I hope you understand I'm never trying to be counterprodocutive. I've golfed bugs out of codebases twice as large. it does take some time and thought, but it does work if you do it right. I only wanted to point out that a tool you already have in your belt can work even here. I hope you don't feel like I'm picking on you. that is never my intention 20:54
Xliff raydiak: I understand that. Thanks for your input. It's just when I come to my p6-GLib projects, golfing is not an easy or quick way to reach a problem. --ll-exception is almost faster. 20:56
20:56 monkey_ joined 20:59 yewscion joined
raydiak for the cases where it works, it definitely is. sometimes the cause is not obvious in the current stack 21:01
did this work when you originally made the change? I wonder what the difference is
21:03 Xliff_ joined, yewscion_ joined 21:04 cnx left, ComplYue[m] left, juanfra left, demostanis[m] left, AlexDaniel left, tyil[m] left, CIAvash left 21:05 juanfra joined, Xliff left, yewscion left, monkey_ left, MasterDuke left 21:06 monkey_ joined, Xliff joined 21:09 demostanis[m] joined, AlexDaniel joined, cnx joined, gabiruh_ joined, ComplYue[m] joined, tyil[m] joined, samcv_ joined, CIAvash joined, andinus` joined 21:11 pierrot_ joined, sivoais_ joined 21:12 broquain1 joined, Ulti_ joined, bartolin_ joined, Xliff_ left, gfldex_ joined 21:14 RandalSchwartz joined, jjatria_ joined, linkable6 left, unicodable6 left, squashable6 left, benchable6 left, tellable6 left, notable6 left, LizBot left, jjatria left, bisectable6 left, quotable6 left, committable6 left, shareable6 left, Ekho left, Ulti left, monkey_ left, samcv left, gfldex left, bartolin left, sivoais left, cognominal_ left, gabiruh left, broquaint left, pierrot left, andinus left, Sgeo_ joined 21:15 jjatria_ is now known as jjatria 21:16 Xliff_ joined 21:17 vrurg_ joined 21:18 Ekho- joined, cognominal joined, samcv joined 21:19 ecocode_ joined, SmokeMachine left, leont left, vrurg left, ecocode left, moon-child left, childlikempress joined 21:20 SmokeMachine joined 21:21 leont joined 21:22 tonyo left, charsbar left, charsbar_ joined, charsbar_ is now known as charsbar 21:23 perlbot_ joined, gfldex_ is now known as gfldex 21:24 perlbot left, jercos left, jdv left, samcv_ left, Xliff left, Sgeo left, jdv joined 21:25 vrurg_ is now known as vrurg, perlbot_ is now known as perlbot, tonyo joined, m6locks left, GoneDomattps is now known as domattps 21:26 jercos joined, m6locks joined 21:30 raydiak left 21:39 monkey_ joined 21:47 monkey_ left 21:49 MasterDuke joined, childlikempress is now known as moon-child
MasterDuke Xliff_: are you using rakudo at HEAD? if so, have you noticed any compilation speedup in the past day or so? 21:51
21:57 sivoais_ left, sivoais joined 22:01 yewscion_ left 22:07 committable6 joined, squashable6 joined, bisectable6 joined, tellable6 joined, benchable6 joined 22:08 quotable6 joined, notable6 joined 22:09 unicodable6 joined, linkable6 joined 22:13 shareable6 joined
Xliff_ MasterDuke: Just compiled at like midnight. 22:58
Or slightly after.
Won't know overall until the singular job completes
That'll be a while
22:59 Xliff_ is now known as Xliff 23:59 linkable6 left, evalable6 left, releasable6 left