Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2006-09-27

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 markstos_ joined perl6
00:09 frederico joined perl6
00:20 beppu_ joined perl6
00:25 Aankh|Clone joined perl6
00:30 nekokak joined perl6
00:43 justatheory joined perl6
00:45 mugwump ajs_work: yeah, that article needs a little work.
00:45 mugwump PGE doesn't make bytecode
00:46 dduncan joined perl6
00:46 mugwump it makes a parse tree.  TGE converts that into an expression tree, then another round to a opcode tree, then it can become bytecode
00:46 mugwump (actually it makes PIR, which is converted by imcc to bytecode, but that's another thing)
01:05 dduncan left perl6
01:05 TreyHarris svn log -v -rBASE:HEAD
01:05 TreyHarris oops
01:17 mako132_ joined perl6
01:29 kanru joined perl6
01:53 mugwump ?eval "foo bar".split(/<ws>/)
01:53 evalbot_r13688 is now known as evalbot_r13694
01:53 evalbot_r13694 ("f", "o", "o", "", "b", "a", "r")
01:53 * mugwump slaps evalbot_r13694
01:53 mugwump ?eval "foo bar".split(/<ws>+/)
01:54 evalbot_r13694 ("f", "o", "o", "", "b", "a", "r")
01:54 mugwump ?eval "foo bar".split(/<[ ]>/)
01:54 evalbot_r13694 ("foo", "bar")
01:56 obra seen audreyt
01:56 jabbot obra: audreyt was seen 7 hours 19 minutes 43 seconds ago
01:58 TreyHarris uh-oh.  I get: *** Cannot parse regex: <ws> *** Error: Error: Can't locate object method "__RUN__" via package "Pugs::Runtime::Match::HsBridge".
01:59 TreyHarris that's when typing that into "./pugs", with my PERL6LIB set to ./blib6/lib
01:59 TreyHarris am I missing something?
02:02 TreyHarris my PUGS_EMBED wasn't set... but I thought pugs embedded perl 5 by default now
02:02 TreyHarris ah, well, i'll rectify that, clean and make...
02:03 Aankhen`` joined perl6
02:11 kyrbe joined perl6
02:27 vel joined perl6
02:34 svnbot6 r13695 | markstos++ | [t/perl5]
02:34 svnbot6 r13695 | markstos++ | Add simple README.
02:37 Qiang joined perl6
02:41 mauke joined perl6
03:05 spoop joined perl6
03:11 mugwump ajs_work: I updated http://en.wikipedia.org/wiki/Parrot_Grammar_Engine
03:11 lambdabot http://tinyurl.com/zvuxw
03:12 TreyHarris nothingmuch: ping?
03:13 TreyHarris oh, i guess it's still 6am there
03:23 kanru joined perl6
03:43 rindolf ingy: here?
03:45 Aankhen`` joined perl6
04:03 Eidolos joined perl6
04:08 Aankh|Clone joined perl6
04:13 TreyHarris anyone on now with Module::Signature clue?
04:18 audreyt I've heard of it...
04:19 TreyHarris lol :)
04:19 TreyHarris i don't know how to fix this.  when i use sudo to install CPAN modules, it always fails.  i just discovered the reason is that when it tries to check the module signature, it runs gnupg, which erroneously says that the module's signature is invalid because my ~/.gnupg belongs to me and not root.  if i chown my .gnupg to root, i can install modules, but then i can't use gnupg as myself...
04:20 audreyt a recent version of CPAN.pm supports sudo only on install
04:20 audreyt so you can invoke cpan as yourself
04:20 audreyt and have it run sudo on your behalf
04:20 audreyt iirc. CPANPLUS also supports that
04:21 audreyt you can also set make_install_make_command to "sudo make install"
04:21 audreyt as well as
04:21 audreyt mbuild_install_build_command
04:21 TreyHarris hmm. i have root's cpan configs set to install into the system, and my person cpan config set to install into my home directory.  so if I do that, Makefile.PL will install into my home directory, but as root, which is the wrong thing in two ways :)
04:22 TreyHarris i guess just "not a supported configuration for Module::Signature" :-)
04:22 audreyt in that case set $GNUPGHOME
04:22 audreyt when invoking cpan as root
04:22 audreyt to point to another set of ~/.gnupg
04:22 audreyt will fix it.
04:22 TreyHarris ahh, that's a good idea.  thanks!
04:22 audreyt np :)
04:29 kyrbe left perl6
04:43 penk joined perl6
05:02 ofer0 joined perl6
05:14 weinig joined perl6
05:29 mjk audreyt: i intend to call "expo" function in power.pl from visual c++,is ok,
05:29 mdiep joined perl6
05:30 mjk but when use "perl -MO=Bytecode,-H,-opower.plc power.pl" compile power.pl to power.plc , vc call "expo" is failed, why?
05:30 mjk can you help me
05:36 ofer0 joined perl6
05:44 svnbot6 r13696 | mugwump++ | Hobble Pugs-Compiler-Perl6 t/regex.t into at least running
06:09 marmic joined perl6
06:21 agentzh joined perl6
06:21 agentzh TimToady: thanks!
06:21 lambdabot agentzh: You have 1 new message. '/msg lambdabot @messages' to read it.
06:23 traecer joined perl6
06:31 iblechbot joined perl6
06:41 frankg joined perl6
06:50 svnbot6 r13697 | audreyt++ | * Fix some incorrect plan lines.
06:50 svnbot6 r13698 | audreyt++ | * "1.5 ~~ 1^..^2" is not for this release; TODO it.
06:59 Southen_ joined perl6
07:02 kastol joined perl6
07:07 rgs joined perl6
07:10 agentzh ?eval say "Hi" when 1;
07:10 evalbot_r13694 is now known as evalbot_r13698
07:10 evalbot_r13698 Error:  Unexpected "when" expecting comment, operator, ":", ",", statement modifier, ";" or end of input
07:11 rgs joined perl6
07:13 kane-xs joined perl6
07:17 zakharyas joined perl6
07:23 xinming joined perl6
07:23 traecer joined perl6
07:28 b00t joined perl6
07:31 drrho joined perl6
07:43 ludan joined perl6
07:45 TreyHarris audreyt: should the prereqs for Pugs::Compiler::Rule be added to pugs' prereqs given in INSTALL?
07:48 TreyHarris oh, wait... Pugs::Compiler::Rule itself is given as a prereq.  but isn't it part of pugs?
07:57 obladoc joined perl6
07:58 agentzh can ``when'' be used as a statement modifier? it seems quite natural.
08:01 Juerd Yes
08:03 TreyHarris Juerd: you can?  that is unspecced.
08:03 TreyHarris I brought this up on p6-l a couple weeks ago
08:03 Juerd Oh
08:04 Juerd I thought it was specced.
08:04 Juerd ?eval given 5 { 3 when 5; 4 }
08:04 evalbot_r13698 Error:  Unexpected "when" expecting comment, operator, statement modifier, ";" or "}"
08:04 Juerd Oh.
08:04 TreyHarris as currently specced, it's not a statement modifier; and even if it were, outside of a given it would cause the enclosing block to be terminated, which i think would not be desired behavior...
08:06 baest joined perl6
08:06 TreyHarris my proposal was that when basically become an alias for "if $_ ~~ " outside of blocks that want given/when breaking behavior (right now, given and CATCH)
08:07 TreyHarris then it could naturally be a statement modifier.
08:33 agentzh *nod*
08:33 elmex joined perl6
08:33 * agentzh is still busy making slides.
08:34 bsb left perl6
08:37 ofer0 joined perl6
08:42 rodi joined perl6
08:46 kane-xs_ audreyt: yeah, CPANPLUS supports that.. we coded that together :) i think CPAN.pm recently ported that feature to support sudo
08:53 TreyHarris joined perl6
09:05 TreyHarris ?eval /msg nickserv identify spitatme
09:05 evalbot_r13698 Error:  Unexpected end of input expecting "\\", variable name or "/"
09:05 TreyHarris oops :-)
09:05 TreyHarris time to change that password :-)
09:06 Juerd spit...at...you...?
09:06 ruoso joined perl6
09:06 Juerd If you wish.
09:06 * Juerd spits :)
09:06 TreyHarris it's memorable ;-)
09:06 * rodi ducks
09:06 buetow joined perl6
09:07 axarob joined perl6
09:10 TreyHarris now that's changed, i can type again. :-) my throw-away passwords are all derivable from an aspect of the service/servername via an algorithm that if i say anything more about would be guessable.  but it results in smashed-together english words of various sorts, sometimes humorously :-)
09:11 TreyHarris keeps me from having to write stuff down but lets me have a different password for each thing i use that wants a password.
09:11 jferrero joined perl6
09:14 TreyHarris ?eval say "\x00ABhi\x00BB"
09:14 evalbot_r13698 OUTPUT[«hi» ] Bool::True
09:16 chris2 joined perl6
09:17 TreyHarris ?eval say "\x7FBD\x91CC\x9808"
09:17 evalbot_r13698 OUTPUT[羽里須 ] Bool::True
09:18 TreyHarris bleh, and i thought i had finally figured out this utf8 irc thing.  the *first* of those three characters displayed...
09:18 * TreyHarris shrugs
09:19 Juerd Awwh
09:20 * Juerd hâş ütf⑧  wørkíñ qùit€ wēłl.
09:21 Juerd s/ñ/ñģ/
09:21 Juerd Perl should complain when you output a text string, by the way
09:21 TreyHarris hâ
09:21 TreyHarris hâ�_ ütf�Q�  wørkíñ qùit�B� w�S�Bl.
09:21 Juerd It seems you have latin1 working :)
09:22 TreyHarris and ç¾½
09:22 TreyHarris isn't that odd.
09:22 Juerd TreyHarris: Depends.
09:22 Juerd TreyHarris: Do you run screen?
09:23 TreyHarris yes.  but my LOCALE is set to en_US.utf8, and I started it with -U
09:23 TreyHarris which i was told was the magic incantation
09:23 Juerd Does that locale *exist*?
09:23 TreyHarris locale -a says it does...
09:23 Juerd Spelled like that?
09:23 Juerd For example, in locales, UTF-8 and utf8 *are* different.
09:24 TreyHarris no, i did that from memory
09:24 TreyHarris ahh... but I have en_US.utf-8.  not UTF-8
09:25 Juerd Next... does irssi know you're utf8'ing now?
09:25 TreyHarris well, i better fix my incorrect casing first...
09:29 TreyHarris joined perl6
09:29 TreyHarris ok... try that message again, Juerd :)
09:30 TreyHarris why shouldn't pugs let me output that, by the way?  or did i misunderstand you?
09:32 * Juerd hâş ütf⑧  wørkíñģ qùit€ wēłl.
09:33 Juerd Because that would assume knowledge of the outside environment, that it doesn't have yet.
09:33 b00t joined perl6
09:33 Juerd Though until we can tell it that we're utf8 ready, this is probably the best way to handle it ;)
09:33 Juerd TreyHarris: How's this?
09:34 TreyHarris i only got the same characters
09:36 Juerd Okay. Does your irssi know about your utf8?
09:36 TreyHarris term_type = utf8
09:36 Juerd /set term_charset utf8
09:36 Juerd Err, UTF-8
09:37 TreyHarris ah... my irssi doesn't have term_charset.  only term_type
09:37 TreyHarris is that my problem?
09:37 Juerd Mine the other way around
09:40 TreyHarris my irssi docs say term_type, but they are inconsistent as to whether they want 'utf-8' or 'utf8' or 'UTF-8', they use all three
09:49 TreyHarris i figured it out, maybe
09:50 TreyHarris looks like I have to give *both* screen -U *and* :utf8 on to each window
09:50 TreyHarris just one or the other isn't sufficient
09:50 mugwump I had to get a trunk irssi for good unicode support
09:50 wolverian that's an old irssi
09:50 mugwump but that was a while ago
09:51 wolverian you should get 0.8.10 for the recode support
09:52 kane-xs_ joined perl6
09:53 TreyHarris wolverian: doesn't matter... maybe irssi isn't playing nice, but just cat'ing a utf8 file was giving me question marks until i did both -U and :utf8
09:53 TreyHarris so it wasn't irssi's fault
09:53 wolverian my locale is UTF-8, so screen follows that automatically
09:53 TreyHarris ah.  just plain "UTF-8"?  my locale -a doesn't have that.
09:54 TreyHarris so I'm using en_US.UTF-8
09:54 wolverian I'm using fi_FI.utf8
09:55 mugwump I'm on en_NZ.UTF-8
09:55 wolverian is that a new hallucinogen?
09:55 * TreyHarris smacks his head
09:56 * mugwump checks PIHKaL
09:56 TreyHarris ahh.  i see now that since i have and ssh involved too, my source machine's locale has "en_US.UTF-8", but my destination machine has "en_US.utf8".
09:56 rodi out of curiousity, what are people using for input methods for characters not present on their respective keyboards?
09:57 wolverian rodi, altgr+various combinations, or scim+anthy for japanese
09:57 Alias_ joined perl6
09:58 TreyHarris it looks like of a) terminal font, b) terminal charset, c) local host LOCALE, d) remote host LOCALE, e) screen run with -U, f) screen window in which irssi is running has :utf8 set, g) irssi has term_type set to UTF-8...
09:58 TreyHarris every one of those has to be explicitly set.  accepting defaults, even if the docs claim it should work, is not sufficient.
09:58 wolverian weird.
09:59 mugwump I wish xkb would let you assign arbitrary unicode chars to key combos
09:59 wolverian I suppose some people might have set screen to not follow locale
09:59 * wolverian shrugs
09:59 TreyHarris ah well.  enough madness for now.  sleep.  thanks for all the help!
09:59 mugwump So I could bind the hotsprings character to something
10:02 TreyHarris ?eval say "\x7FBD\x91CC\x9808"
10:02 evalbot_r13698 OUTPUT[羽里須 ] Bool::True
10:02 TreyHarris ok, mugwump, you're right, irssi isn't helping either.  (only the first character displays.)  but that displays on my other screens, so I'm set I think :-)
10:03 * TreyHarris Zzzzzz
10:04 Alias_ I met the oddest person today
10:04 Alias_ Statistically at least
10:04 Alias_ A Good Female Australian Muslim Programmer with a Functional language fetish
10:05 Juerd mugwump: It does that
10:05 Alias_ The odds of all those combining must be exceptionally rare
10:05 Juerd mugwump: find Compose stuff on your computer
10:05 mugwump Juerd: but any arbitrary codepoint?
10:05 Juerd mugwump: AFAIK, yes
10:06 mugwump hmm
10:06 mugwump it uses words like "EuroSign", "dead_ogonek", etc
10:07 Juerd And literals
10:08 Juerd dead_ogonek are keyboard symbols
10:08 Juerd Not result character
10:08 Juerd The resulting character is after the colon, in quotes
10:10 mugwump do you have an example?
10:10 Juerd Plenty in my Compose files.
10:10 Juerd But I can't copy/paste
10:11 Juerd And since you need to locate the right file anyway, I'm not going to type a line :)
10:11 mugwump is there a file I should be looking in in the X.org dist (or as typically installed at...) ?
10:14 Juerd /usr/share/X11/locale
10:14 Juerd find Compose there
10:14 mugwump ah yes... cool
10:15 mugwump oh wow, lots already there in the en_US.UTF-8
10:15 Juerd Yea :)
10:16 mugwump Darn, not enough free keys on this laptop
10:16 * mugwump re-binds his menu key to Multi_key
10:16 Juerd I use the menu key too
10:17 Juerd On my laptop, which has no winkeys, the right alt.
10:19 mugwump c¤øl
10:19 rodi left perl6
10:20 rodi joined perl6
10:21 Juerd V€ry ¢ôôł
10:22 mugwump At last – relief for typographers — Said the unicode-endowed page
10:22 frankg joined perl6
10:23 mugwump hmm, actually I think my en dashes and em dashes are getting recoded
10:24 Juerd I saw en dashes.
10:25 Juerd Though on a text terminal, they're the same perhaps
10:25 Juerd After all, m is as wide as n
10:25 Juerd I have unicode working in Mutt since yesterday
10:25 Juerd And already I've received two cyrillic spams.
10:25 Juerd How did they knowL
10:26 Juerd s/L/?/
10:30 mugwump wǒde hànyǔ shì awful, but at least I can type it in pinyin easily armed with these key combinations
10:36 Juerd H₂O
10:36 Juerd Hehe² :)
10:40 iblechbot joined perl6
10:57 markstos_ joined perl6
11:09 ofer0 joined perl6
11:09 axarob left perl6
11:22 Alias_ left perl6
11:23 foo\ joined perl6
11:28 KingDiamond joined perl6
11:34 xinming joined perl6
12:01 shizu joined perl6
12:06 rindolf joined perl6
12:06 rindolf Hi all.
12:06 rindolf ingy: ping.
12:07 * [particle] notes it's 5am ingy time
12:07 [particle] ...but it's 5am particle time, too
12:07 rindolf [particle]: I see.
12:09 rindolf [particle]: are you up early or up late?
12:16 [particle] i'm up early.
12:16 [particle] got up at 4:30, 1.5hrs early :(
12:38 nothingmuch_ joined perl6
12:58 fglock joined perl6
13:00 ofer1 joined perl6
13:03 elmex joined perl6
13:27 rodi joined perl6
13:27 crem joined perl6
13:29 svnbot6 r13699 | fglock++ | Pugs-Compiler-Rule-0.19 in CPAN
13:34 dakkar joined perl6
13:36 c6rbon_ joined perl6
13:43 vel joined perl6
13:43 chris2 joined perl6
13:48 cjeris joined perl6
14:02 c6rbon joined perl6
14:12 wolverian joined perl6
14:14 plural joined perl6
14:15 audreyt joined perl6
14:15 chip joined perl6
14:15 pmurias joined perl6
14:15 PerlJam joined perl6
14:15 xinming_ joined perl6
14:15 leo_ joined perl6
14:15 pmichaud_ joined perl6
14:15 webmind joined perl6
14:15 Juerd joined perl6
14:30 lanny joined perl6
14:34 lanny Hmm.  Why are the Synopses generated from the same revision different on different websites?  Re: http://feather.perl6.nl/syn/S29.html and http://m19s28.vlinux.de/iblech/stuff/pugs-smokes/pugs-smoke-6.2.12-r13695-linux-normal--1159330463-2215--12745-12088-657-1510-833-10--b22f10acb37fc685a83f7261a3be11dc-synopses/S29.html
14:34 lambdabot Title: S29
14:37 Gothmog__ joined perl6
14:38 kanru joined perl6
14:43 rodi lanny: AFAIK, S29 lives in the pugs repo, and those two are different pugs revs.  I think the other synopses mostly live in the syn repository.
14:44 rodi It's weird.  I think if you check out the synopses directly, S29 is just a few-line placeholder.
14:44 lanny Even granted that TimToady changed S29 a few days ago.  The pages should be in sync
14:45 lanny And it's only a lead-up to me asking why the smartlinks I put in place for t/builtins/math/*.t aren't showing up.  :)
14:46 rodi haha, gotcha.  you make a good point, though- last touch on that was r13589.  Well, I'm out of ideas.  That didn't take me long ;)
14:49 lanny Even scarier when I look at svn.openfoundry.org/pugs/docs/Perl6/Spec/ I don't find the files at 13699 I do in my co of 13699
14:51 lanny Ah.  Well that explains that.  They aren't under svk control in my co.  :/
14:57 lanny Ok.  smartlinks.pl hits svn.perl6.org and populates the missing ones.  But svn.openfoundry.org/pugs/docs/Perl6/Spec/Functions.pod is *the* S29 so the original observation is still correct; different S29s on different websites
14:58 lanny and broken smartlinks
15:06 Odin-LAP joined perl6
15:17 weinig joined perl6
15:29 putter joined perl6
15:31 lanny @tell agentzh Not sure if this should be directed to you but Synopsis seem to be different between latest smokes on m19s28.vlinux.de and feather.
15:31 lambdabot Consider it noted.
15:33 putter @tell audreyt r13604 "* splatty_in_caller_position.t: It's f(|@args), not f(*@args)." looks like it deleted 15 t/regex test files.  was that intentional?
15:33 lambdabot Consider it noted.
15:34 justatheory joined perl6
15:41 Aankhen`` joined perl6
15:46 buubot joined perl6
15:48 lanny quit
15:49 Gothmog__ is now known as Gothmog_
15:49 elmex joined perl6
15:50 buubot joined perl6
16:03 putter fglock: ping?
16:25 Psyche^ joined perl6
16:26 hexmode joined perl6
16:32 vel joined perl6
16:39 Psyche^ is now known as Patterner
17:13 TreyHarris @tell lanny the reason S29 is not in sync is that S29 is still considered "non-stable" and under community control so anyone with pugs commit can edit it, and there's no maintainer or gatekeeper.  the other synopses have maintainers (with TimToady as uber-maintainer).  so they're in two different svn repos.  the text link is to deal with not being able to symlink from one svn repo to another portably
17:13 lambdabot Consider it noted.
17:15 Ser_ joined perl6
17:26 hexmode joined perl6
17:27 fglock joined perl6
17:27 fglock joined perl6
17:29 fglock putter: pong
17:29 mdiep joined perl6
17:31 ajs_work @tell lanny more generally, S29 in the pugs repo is the "working copy" and will eventually be/replace the authoritative version once it reaches stability.
17:31 lambdabot Consider it noted.
17:33 buetow joined perl6
17:33 putter hi fglock
17:33 putter repeat patch inbound...
17:33 putter err, range patch...
17:34 putter not the whole thing.
17:34 putter just the logic plumbing
17:35 putter 13701.
17:36 fglock putter: cool - if you implement the basics in PCR, then v6.pm can plug "real code" into it
17:36 fglock (real code = iterators)
17:37 putter it was kind of neat that with goto passing @_ along, much of the range logic doesn't need to understand the runtime's calling conventions.
17:37 putter the parts that do, I left for you, as I really don't. ;)
17:39 TimToady I just made :ratchet autoanchor on both ends unless overridden by :p or :c (on the assumption that subrule calls supply :p)
17:40 TreyHarris how is <ws> defined?  I'm still trying to understand
17:40 TimToady so token { \d+ } is equiv to m/ ^ \d+: $ /
17:40 TreyHarris ?eval "foo bar".split(/<ws>/)
17:41 TimToady that's not what <ws> i sfor.
17:41 TreyHarris oh, evalbot's down
17:41 TimToady <ws> is the optional whitespace matcher used by :sigspace.
17:41 TreyHarris TimToady: yes, I know, but somebody else typed it and i'm trying to understand it :)
17:41 TimToady so it can be the null string between non-word chars.
17:41 TreyHarris it's apparently the null string between word characters, too
17:41 TimToady well, in that case it should work.
17:42 TimToady that's a bug
17:42 fglock TimToady: do you have a test case for <ws> ?
17:42 TreyHarris ah, ok
17:42 pasteling "putter" at 66.30.119.55 pasted "svnbot is apparently on vacation... r13701 commit message." (12 lines, 818B) at http://sial.org/pbot/20008
17:43 TimToady you mean like the one TreyHarris just typed?  I think that's already in the test suite somewhere.
17:43 putter fglock: just let me know if you have any questions or issues.  cheers.
17:43 TimToady well, I guess not...
17:44 fglock putter: looking - thanks!
17:44 putter :)
17:44 putter lunch.  bbl &
17:44 TreyHarris i couldn't find <ws> tested directly in the test suite
17:45 TreyHarris it's used a couple places... but not tested
17:45 TreyHarris unless m:p5 /<ws>/ is not a valid way to look for it
17:47 TimToady I did a recursive grep for "split.*ws" but didn't find it.
17:49 TimToady a lot of them are <?ws>
17:49 rodi I asked a question about this yesterday- doc/Perl6/Overview/Rule.pod has a number of built-in assertions listed that are not in S05.  Some have tests, some don't.  I'd happily do the bookeeping of adding some tests and links if I had a clearer idea of which list is "official-er"
17:51 TimToady S05 is officialer.  I haven't even looked at Overview/Rule.  But if Overview points out holes in S05 we should plug 'em.
17:51 PerlJam rodi: which does docs/Perl6/Overview/Rule.pod have that aren't in S05?
17:53 rodi PerlJam: the ones I've noticed: <dot>, <lt>, <gt>, <fail>, <null>, <ident>, <self>, and <'...'>.  <ident> is used in S05, if I recall correctly, but not defined.
17:53 TreyHarris ajs_work: reading your most recent mail.  under your proposal, how would one specify the proto of a multi that takes several (but not all) types for its args?  conjunct them?
17:54 PerlJam <null> is in S05
17:54 rodi sorry, <null> is defined, just not in the list with <before pattern>, <after pattern>, etc.
17:54 rodi sic.
17:55 PerlJam some of those others are the same.
17:55 TreyHarris er... for example
17:55 TreyHarris ajs_work: por ejemplo, delete
17:56 fglock does <null> captures? do I have to use <?null> instead?
17:57 PerlJam fglock: yes and yes.
17:57 PerlJam though capturing <null> may not hurt :)
17:57 fglock how about <before pattern> and <?before ...>
17:57 TreyHarris "our List proto delete(Array|Hash $container: *@indices)"?
17:58 rindolf joined perl6
17:58 rindolf Hi all.
17:58 rindolf ingy: here?
17:58 rodi fglock: I don't think <before pattern> and <after pattern> currently capture in pugs.
17:58 PerlJam fglock: as I understand things <assertion> always "captures" what is matched into the Match object and <?assertion> does not.
17:59 TreyHarris perhaps Array and Hash do "Indexed" or some such, then you can use "our List proto delete(Indexed $container: *@indices)"...
17:59 PerlJam fglock: though I believe that it is captured lazily if such a thing makes sense
17:59 PerlJam (i.e., it keeps track of .from and .to but doesn't actually copy the part of the string that matched until you ask for it
18:00 PerlJam )
18:00 TreyHarris though that means that Indexed forces the requirement of implementing .delete on its... erm.  "implementors"?  what's the word for a user of a role?
18:00 fglock PerlJam: lazy captures are a problem - if it captures once, it returns a match, but if it captures many, it returns an array - it is difficult to deal with the return value
18:00 [particle] why would a *zero-width* assertion capture?
18:01 fglock PerlJam - oh, the string is never copied until you ask for it
18:01 [particle] ...and capturing <null> will likely be slower than non-capturing <?null> due to implementation
18:01 fglock [particle]: yes, I'm asking if non-<?...> always captures or not
18:02 TreyHarris is there a reason that @array<0> souldn't work?
18:02 PerlJam It's a good rule that <X> always captures and <?X> always does not.
18:03 PerlJam (though from an implementation perspective, I kind of wish it were the other way)
18:03 PerlJam but that's just me
18:05 [particle] <before foo> returns a zero-width Match object in PGE
18:05 [particle] $.from and $.to are null
18:05 [particle] ...but $.pos is set
18:14 weinig is now known as weinig|away
18:15 fglock putter: not enough coffee - I can't even understand my own code right now
18:16 bernhard joined perl6
18:27 putter back
18:27 putter fglock: :)
18:44 fglock back in few mins &
19:08 amnesiac joined perl6
19:16 rashakil joined perl6
19:23 TimToady_ joined perl6
19:29 fglock putter: in the current logic, you need to initialize the state, with something like: my @state = $_[1] ? @{$_[1]} : [ 0 ];  # current count
19:33 fglock hmm - it's starting to make sense to me - adding a match counter non_greedy_plus() may be enough
19:42 nothingmuch joined perl6
19:43 nothingmuch yo yo yo homies
19:43 fglock nothingmuch: yo-yo
19:44 nothingmuch yo man you need to show some rezpect or imma get upset
19:44 nothingmuch you know what i'm sayin?
19:45 fglock nah
19:47 pasteling "fglock" at 200.17.89.34 pasted "putter: ranges implementation under way" (16 lines, 478B) at http://sial.org/pbot/20011
19:47 fglock high level functions are scary
20:25 Daveman joined perl6
20:26 kastol joined perl6
20:32 timbunce joined perl6
20:40 arguile joined perl6
20:43 timbunce joined perl6
20:49 fglock putter: r13702 - non_greedy_plus() now accepts (min, max) - added some tests; no emitter side yet
20:51 fglock changing this to accept an iterator is straightforward (given enough coffee)
20:52 justatheory joined perl6
21:00 nothingmuch joined perl6
21:00 pasteling "fglock" at 200.17.89.34 pasted "backtracking non-greedy range quantifier - working" (16 lines, 549B) at http://sial.org/pbot/20012
21:27 Limbic_Region joined perl6
21:29 fglock left perl6
21:32 nothingmuch audreyt: do you have a recursive ddrescue script?
21:32 nothingmuch my HD is occasionally blocking the USB/scsi bridge and i need to powercycle and remount it
21:32 nothingmuch so ddrescue would be *very* useful
21:37 hexmode joined perl6
21:40 araujo joined perl6
21:47 avar joined perl6
21:48 lanny joined perl6
21:51 lanny Admit that I have messages!
21:51 lambdabot lanny: You have 2 new messages. '/msg lambdabot @messages' to read them.
21:51 lanny lambdabot: @messages
21:51 lambdabot TreyHarris said 4h 38m 23s ago: the reason S29 is not in sync is that S29 is still considered "non-stable" and under community control so anyone with pugs commit can edit it, and there's no maintainer
21:51 lambdabot or gatekeeper. the other synopses have maintainers (with TimToady as uber-maintainer). so they're in two different svn repos. the text link is to deal with not being able to symlink from one svn
21:51 lambdabot repo to another portably
21:51 lambdabot ajs_work said 4h 20m 44s ago: more generally, S29 in the pugs repo is the "working copy" and will eventually be/replace the authoritative version once it reaches stability.
21:52 Limbic_Region lanny amdit that you knew you had messages because you read the irc logs prior to logging in!
22:13 jferrero joined perl6
22:29 LimbicRegion joined perl6
22:35 LimbicRegion is now known as Limbic_Region
22:40 nothingmuch joined perl6
23:01 kastol joined perl6
23:24 justatheory joined perl6
23:24 markstos_ joined perl6
23:51 putter joined perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo