perl6-projects.org/ | nopaste: sial.org/pbot/perl6 | evalbot: 'perl6: say 3;' | irclog: irc.pugscode.org/
Set by mncharity on 25 March 2009.
s1n anyone wanna be a guinee pig for my NLP project? 03:35
wayland76 Can you do neuro-linguistic programming through IRC? :) 04:00
s1n wayland76: not sure exactly what you're asking there, but i need an evaluator or two, which can be done via IRC 04:08
oh, NLP = natural language processing
wayland76 Ah, ok :)
what does being a guinea pig involve in this case? 04:09
s1n evaluating generated output
wayland76 evaluating based on what criteria? Grammaticalness? Artistic merit? 04:10
s1n i have a set of weighted grading system i devised for this experiment 04:10
mncharity @tell masak Thanks for the commit! Please let me know if you encounter other difficulties, or if I can be of any help. 04:39
lambdabot Consider it noted.
jnthn oh morning 06:51
Matt-W oh hai jnthn 06:55
wayland76 Ooh, people :) 07:10
Matt-W yes people 07:18
well, me
and maybe jnthn's still hee
here*
wayland76 Well, you see, skids bets that dalek is going to be the biggest talker "tomorrow" 08:04
eiro hello 08:33
Matt-W hello
bacek hi there 10:14
how is Oslo hackatron going?
ihrd hi there 10:29
I use action for grammar tokens, and find out something strange 10:30
If I use anything with regexp inside action method 10:31
like .subst(/\w/, 'Y'); 10:32
make do not work
I think this happens because of make use $/ 10:33
can I change this somehow? can I specify make to use other variable instead?
wayland76 Well, you could try saving $/ into another variable, and restoring it before the end of the action 10:36
ihrd wayland76: thank you, it is works 10:47
btw it looks ugly
I have no possibility to change C<make> behavior? 10:49
wayland76 I've had a better idea 10:50
Use a temp scoped variable
ie. at the start, write "temp $/"
...and it should fix itself when you exit the scope 10:51
not sure if it's implemented, but it's in S02
ihrd but I have make and another match in the same scope 10:52
wayland76 Well, I don't understand P6 grammars properly yet :) 10:54
Matt-W I think there will be a better mechanism for that sort of thing later 10:55
but we don't have one yet
(that I'm aware of)
Matt-W seems to me that $/.ast should be assignable, that'd do the same thing as make does... 10:55
and then you could do it to other objects 10:56
or maybe $match.make($thing)
this is just groundless speculation though
ihrd Matt-W: yes, I was expected something like this 10:57
wayland76: temp var NYI 10:58
Matt-W But there will be something better
If only because I'll be screaming my head off until it appears :)
wayland76 ok
wayland76 practises doing that too. Aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaahgh! 10:59
pmurias ihrd: $/ is contextual so if you the other match in a nested lexical scope it shouldn't cause problems 11:06
ihrd pmurias: yes, and I realy have problems 11:07
ihrd I have some check, like $exp ~~ /^ \s* $/ in action method, and now I should work ugly workaround to use make after 11:10
pmurias ihrd: if you use do { $exp ~~ /^ \s* $/ } the $/ shouldn't be contaminated 11:13
ihrd but it is 11:14
ah, sorry
you are right, I will try
pmurias: yes, do this in block is better workaround 11:16
thank you 11:17
eiro pastebin.com/f4df057e4 15:06
buubot eiro: The paste f4df057e4 has been copied to erxz.com/pb/17070
eiro this gives me <this a> 15:07
i expected <a> ... any idea ? (i take my exemples from www.programmersheaven.com/2/Perl6-F...egex#regex which can be outdated) 15:08
wayland76 rakudo: my %k = < this is a good >; for %k.keys { /a/ or next; .say }; 15:10
p6eval rakudo a4535c: OUTPUT«this␤a␤»
Tene pmichaud: Okay, that's pretty cool... defining a sub named !sub_trait_lol makes "sub foo is lol { ... }" work properly 15:11
eiro that's what i said, wayland76 :)
Tene pmichaud: but I can't define tha sub without some hackery...
wayland76 rakudo: my %k = < this is a good >; for %k.keys { /a/ or next; .say; say /a/; }; 15:13
p6eval rakudo a4535c: OUTPUT«this␤_block42␤a␤_block42␤»
wayland76 rakudo: my %k = < this is a good >; for %k.keys { /a/ or next; .say; say((/a/)); };
p6eval rakudo a4535c: OUTPUT«this␤_block42␤a␤_block42␤» 15:14
Tene wayland76: mentioning a regex isn't the same as matching against it in Perl 6
that's just testing the truthiness of the regex /a/
[particle]- yep
wayland76 I'm trying to print a boolean :)
[particle]- you need to smartmatch
wayland76 oh
Tene wayland76: you'r eprinting a regex, not a boolean.
wayland76 rakudo: my %k = < this is a good >; for %k.keys { /a/ or next; .say; say( $_ ~~ /a/); };
p6eval rakudo a4535c: OUTPUT«this␤␤a␤a␤»
eiro ohhh ... 15:15
Tene rakudo: for <this a> { next unless $_ ~~ /a/; .say }
p6eval rakudo a4535c: OUTPUT«a␤»
eiro that's sad :(
wayland76 rakudo: my %k = < this is a good >; for %k.keys { $_ ~~ /a/ or next; .say; say( $_ ~~ /a/); };
p6eval rakudo a4535c: OUTPUT«a␤a␤»
Tene you know, I wonder if a statement-level version of when would be a good idea
next when /a/
[particle]- except you'll also want the reverse 15:16
Tene when't
;)
[particle]- well, it's legal perl 6, if not proper english :)
Tene when not /a/ should work...
Hm, it doesn't in rakudo... 15:17
ah, nm
I'm no longer convinced it should work.
eiro ~~ could be unary ? 15:18
next if ~~/a/
Tene TimToady: what do you think about statement-modifier 'when'?
wayland76 @seen TimToady 15:21
lambdabot TimToady is in #perl6. I don't know when TimToady last spoke.
literal I recall that not making sence because 'when' leaves the block after it's executed 15:22
Tene literal: and this would do the same.
literal oh, ok
Tene next when /a/; seems nicer than when /a/ { next }
literal why not just "return unless /a/" ? 15:23
Tene because 'next' will start up the next iteration of the loop
[particle]- next unless $_ ~~ /a/;
Tene oh, and just mentioning a regex in Perl 6 doesn't match it against $_
Sure, that works fine. Just a little wordier than might be ideal. 15:24
wayland76 when /a/ { next; } # is good enough -- at least we don't have to refer to the context variable
Tene since automatically matching against $_ has been removed, it might be nice to have a keyword for a common form of it.
wayland76 rakudo: my %k = < this is a good >; for %k.keys { when !/a/ { next; } .say; }; 15:25
p6eval rakudo a4535c: OUTPUT«Statement not terminated properly at line 1, near ".say; };"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: my %k = < this is a good >; for %k.keys { when /a/ { next; } .say; };
p6eval rakudo a4535c: OUTPUT«Statement not terminated properly at line 1, near ".say; };"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤» 15:26
Tene wayland76: };
semicolon after your closing }
wayland76 rakudo: my %k = < this is a good >; for %k.keys { when /a/ { next; }; .say; };
p6eval rakudo a4535c: OUTPUT«this␤»
wayland76 rakudo: my %k = < this is a good >; for %k.keys { when !/a/ { next; }; .say; };
p6eval rakudo a4535c: OUTPUT«this␤a␤»
wayland76 Is there a way to do what I mean in my last example? 15:27
Tene when /a/ { .say };
wayland76 Well, not quite what I want, but I guess I could live with smartmatch 15:30
Tene std: sub 'quoted_subname' { say 'omg i am quoted' } 15:39
p6eval std 26238: OUTPUT«##### PARSE FAILED #####␤Malformed routine at /tmp/M3L9Ez2s1W line 1:␤------> sub 'quoted_subname' { say 'omg i am quoted'␤ expecting any of:␤ blockoid␤ name␤ routine_def␤ terminator␤ trait␤FAILED 00:02 34m␤»
Tene hmm...
std: infix:<kindaquote> { say 'lol' } 15:40
p6eval std 26238: OUTPUT«ok 00:02 35m␤»
Tene std: sub infix:<kindaquote> { say 'lol' }
p6eval std 26238: OUTPUT«ok 00:02 35m␤»
Tene std: sub <kindaquote> { say 'lol' }
p6eval std 26238: OUTPUT«##### PARSE FAILED #####␤Malformed routine at /tmp/pMlRl1gUFA line 1:␤------> sub <kindaquote> { say 'lol' }␤ expecting any of:␤ blockoid␤ name␤ routine_def␤ terminator␤ trait␤FAILED 00:02 34m␤»
mikehh rakudo (a4535c1) builds on parrot r38157 - make test PASS / make spectest PASS - t/spec/S09-typed-arrays/hashes.rakudo - TODO passed: 6-7 16:07
Ubuntu Intrepid i386 - in fact ./perl6 t/spec/S09-typed-arrays/hashes.t ok all inc. 2 TODO and 1 SKIP 16:08
have we any feedback from other builds on this? 16:09
kadaver what is audrey Tang's irc name? 16:15
[particle]- audreyt 16:16
...and she hasn't been around for a long time.
kadaver is she still working on it? 16:17
[particle]- it?
kadaver Pugs I mean. and is Pugs the main Perl 6 implementation+
[particle]- no, and no.
pugs hasn't been under active development for a few years.
rakudo is the perl 6 implementation with the most passing tests, and most active development 16:18
rakudo.org/how-to-get-rakudo
pmichaud: ping 16:19
wayland76 The other ones are smop and elf
smop is also actively developed by multiple developers 16:20
[particle]- is that all there are, these days?
wayland76 There may be another, but I forget
[particle]- perl6-projects.org/
wayland76 I was going to write more, but try [particle]-'s link instead
pyrimidine wasn't audreyt waiting on a newer version of ghc for some reason before taking up Pugs again? 16:21
[particle]- yes. has ghc 6.10 come out?
pyrimidine yes: haskell.org/ghc/
[particle]- well then, maybe she's waiting on other things, or has no tuits. 16:22
pyrimidine my thought as well
[particle]- seems ghc 6.10 isn't a blocker anymore
wayland76 Probably waiting for enlightenment DR17 :)
kadaver 6.10.2 now 16:26
tuits? 16:27
[particle]- www.perlfoundation.org/perl5/index.cgi?tuit 16:28
pyrimidine re: pugs, I worry more about how much has changed with the spec. 16:30
anyone know how much of the test suite it passes? 16:31
(at one point I thought it was ~16000 or so)
[particle]- i haven't built pugs in a very long time 16:32
pyrimidine but that's before moritz's overhaul
yes, same here
[particle]- maybe someone on feather could run pugs against the spec tests
i bet there are a ton of parse errors now
pyrimidine Agreed. And, seeing as rakudo has passed 10000 it appears to be driving the spec now. 16:34
(well. has been for a while actually)
but maybe I'm wrong? 16:35
kadaver is Larry Wall not working on Perl 6?
pyrimidine I would say he is 16:36
irclog.perlgeek.de/search.pl?channe...ady&q=
(note the nick) 16:37
[particle]- larry is hard at work on perl 6 16:39
ruoso I didn't know the irclog supported nick filter... 16:43
kadaver well audreyt is getting really beautiful these days 16:45
[particle]- and tomorrow is audrey's birthday 16:52
wayland76 I wonder if geeks are more likely to look better as they get older than other people are 17:06
I know I look better at 30-something than I did at, say 23
skids Maybe it's the anti-oxidant effect of caffeine? :-) 17:11
wayland76 Well, I don't do tea or coffee either 17:27
For me, it's because I took the time to study appearance like a geek would. Research historical styles, learn the difference between fashion and style, and develop a still-unique style that actually looks alright 17:29
I met a friend that I rarely see, and he said I now look like a jazz musician 17:30
wayland76 I'm wondering if other geeks do something similar sometime between the ages of 25 and 35 17:32
PerlJam only the insane ones. 17:33
wayland76 Well, maybe :) 17:34
I didn't mean that we all end up looking like jazz musicians, just that we realise that appearance can be studied from a geek point of view
PerlJam The phenomenon you're referring to is just that geeks are typically late (if ever) bloomers, in general.
wayland76 you mean, late or early, depending on topic and interest? 17:38
PerlJam geeks are typically intellectually advanced as compared with their peers, but often physically/emotionally retarded by comparison. 17:39
wayland76 Oh, of course :). I know that, I was just thinking too ... specifically. Maybe I shouldn't think out loud at 3am :) 17:45
PhatEddy rakudo: say ("a" .. "z")[0] 18:53
p6eval rakudo a4535c: OUTPUT«a␤»
PhatEddy Anyone around to close RT 61848 if such is merited? (rt.perl.org/rt3/Public/Bug/Display....?id=61848) 18:54
Tene PhatEddy: I've resolved it. Thanks. 19:00
PerlJam Was there also a test for it?
PhatEddy I found one at the end of t/spec/S02-builtin_data_types/range.t I think ... 19:08
omega I'm trying to get november (and as a dep html-template) to run, but I get the following error trying to make html-template: pastie.org/449183 21:51
wayland76 Don't forget to also ask on #november-wiki
omega ahh
wayland76 (although most of those people hang out here too)
omega and I guess they will send me to #text-escape ? :p 21:52
wayland76 ...Or to buy a left-handed screwdriver :)
omega something like that :p 21:53
ohh well, I guess I can ask at the talk at npw tomorrow :p 21:55
just wanted to read up and look at it
cspencer perl6: my $s = "abcde"; my $p = "XxXxX"; say $s.samecase($p)
p6eval pugs: OUTPUT«*** No such method in class Str: "&samecase"␤ at /tmp/Juz6UkqIpn line 1, column 39 - line 2, column 1␤» 21:56
..rakudo a4535c: OUTPUT«Method 'samecase' not found for invocant of class 'Str'␤current instr.: 'parrot;P6metaclass;dispatch' pc 644 (src/classes/ClassHOW.pir:163)␤»
..elf 26238: OUTPUT«Can't locate object method "samecase" via package "abcde" (perhaps you forgot to load "abcde"?) at (eval 125) line 5.␤ at ./elf_h line 5881␤»
wayland76 best thing to do. The IRC channels are relatively dead with NPW on
cspencer perl6: my $s = "abcde"; my $p = "XxXxX"; say samecase($s, $p)
p6eval rakudo a4535c: OUTPUT«Could not find non-existent sub samecase␤current instr.: '_block14' pc 101 (EVAL_17:52)␤»
..pugs: OUTPUT«*** No such subroutine: "&samecase"␤ at /tmp/uOPPIU6i6l line 1, column 39 - line 2, column 1␤»
..elf 26238: OUTPUT«Undefined subroutine &GLOBAL::samecase called at (eval 125) line 5.␤ at ./elf_h line 5881␤»
jnthn I think there wasn't wifi at the NPW. 21:59
Can't say I've missed it though.
omega there was some sometimes 22:03
jnthn ah 22:04
I didn't notice it the times I had my laptop open and didn't look much the other times...

Powered by Cro and the Raku® Programming Language.

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

Thank you!