Camelia, the Perl 6 bug

IRC log for #perl6, 2014-07-24

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

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

All times shown according to UTC.

Time Nick Message
00:03 BenGoldberg joined #perl6
00:06 rurban1 joined #perl6
00:10 Mouq m: ∅
00:10 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/EFrtRZey8nâ�¤Bogus statementâ�¤at /tmp/EFrtRZey8n:1â�¤------> [32m<BOL>[33mâ��[31m∅[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix or meta-prefixâ�¤Â»
00:10 Mouq std: ∅
00:10 camelia std 0f2049c: OUTPUT«[31m===[0mSORRY![31m===[0m�Bogus statement at /tmp/gvyk70ZGrh line 1:�------> [32m<BOL>[33m�[31m∅[0m�Parse failed�FAILED 00:01 122m�»
00:10 * Mouq is confused
00:18 BenGoldberg Totally bogus! :)
00:19 BenGoldberg m: ∅.say
00:19 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/a7XqidXTEDâ�¤Bogus statementâ�¤at /tmp/a7XqidXTED:1â�¤------> [32m<BOL>[33mâ��[31m∅.say[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix or meta-pr…»
00:19 BenGoldberg m: say (∅)
00:19 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/hNkm3_e8i6â�¤Bogus statementâ�¤at /tmp/hNkm3_e8i6:1â�¤------> [32msay ([33mâ��[31m∅)[0mâ�¤    expecting any of:â�¤        semicolon listâ�¤        prefix or termâ�¤        prefix or meta-prefi…»
00:19 BenGoldberg m: say q∅bogus!∅
00:19 camelia rakudo-moar b20535: OUTPUT«bogus!␤»
00:20 atroxaper joined #perl6
00:22 TimToady m: constant ∅ = set; say ∅
00:22 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/xYaaK0ZhA4â�¤Missing initializer on constant declarationâ�¤at /tmp/xYaaK0ZhA4:1â�¤------> [32mconstant [33mâ��[31m∅ = set; say ∅[0mâ�¤    expecting any of:â�¤        statement listâ�¤       …»
00:23 Psyche^ joined #perl6
00:23 TimToady m: constant term:<∅> = set; say ∅
00:23 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/0Fv7nKlfNaâ�¤Two terms in a rowâ�¤at /tmp/0Fv7nKlfNa:1â�¤------> [32mconstant term:<∅> = set; say [33mâ��[31m∅[0mâ�¤    expecting any of:â�¤        argument listâ�¤        prefix or termâ�¤ …»
00:23 TimToady m: my \term:<∅> = set; say ∅
00:23 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/fb5Ud0qk3sâ�¤Term definition requires an initializerâ�¤at /tmp/fb5Ud0qk3s:1â�¤------> [32mmy \term[33mâ��[31m:<∅> = set; say ∅[0mâ�¤    expecting any of:â�¤        scoped declaratorâ�¤Â»
00:23 TimToady std: my \term:<∅> = set; say ∅
00:24 camelia std 0f2049c: OUTPUT«[31m===[0mSORRY![31m===[0m�The 'set' listop may not be called without arguments (please use () or whitespace to clarify) at /tmp/crwVzpWW6Y line 1:�------> [32mmy \term:<∅> = set[33m�[31m; say ∅[0m�Check failed�FAILED 00:01 126m�»
00:24 TimToady std: my \term:<∅> = set(); say ∅
00:24 camelia std 0f2049c: OUTPUT«ok 00:01 126m␤»
00:24 TimToady m: my \term:<∅> = set(); say ∅
00:24 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/0ffzozuPIIâ�¤Term definition requires an initializerâ�¤at /tmp/0ffzozuPII:1â�¤------> [32mmy \term[33mâ��[31m:<∅> = set(); say ∅[0mâ�¤    expecting any of:â�¤        scoped declaratorâ�¤Â»
00:24 TimToady std: constant term:<∅> = set(); say ∅
00:24 camelia std 0f2049c: OUTPUT«ok 00:01 126m␤»
00:25 TimToady m: sub term:<∅> { set() }; say ∅
00:25 camelia rakudo-moar b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/1a5r9xNym5â�¤Two terms in a rowâ�¤at /tmp/1a5r9xNym5:1â�¤------> [32msub term:<∅> { set() }; say [33mâ��[31m∅[0mâ�¤    expecting any of:â�¤        argument listâ�¤        prefix or termâ�¤  â€¦Â»
00:26 TimToady when did that stop working?
00:27 TimToady or contrariwise, why is it complaining about feeding a term to say?!?
00:28 TimToady m: sub term:<∅> { set() }; say 42
00:28 camelia rakudo-moar b20535: OUTPUT«42␤»
00:28 TimToady m: sub term:<es> { set() }; say es
00:28 camelia rakudo-moar b20535: OUTPUT«set()␤»
00:28 TimToady seems like a unicode feyl
00:30 TimToady p6: sub term:<∅> { set() }; say ∅
00:30 camelia niecza v24-109-g48a8de3: OUTPUT«set()␤»
00:30 camelia ..rakudo-jvm b20535: OUTPUT«Unhandled exception: java.lang.RuntimeException: Missing or wrong version of dependency 'gen/jvm/stage2/QRegex.nqp'␤  in  (src/Perl6/Pod.nqp)␤  in  (src/vm/jvm/ModuleLoader.nqp:78)␤  in load_module (src/vm/jvm/ModuleLoader.nqp:60)␤  in  (src/Perl6/A…»
00:30 camelia ..rakudo-{parrot,moar} b20535: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Two terms in a rowâ�¤at /tmp/tmpfile:1â�¤------> [32msub term:<∅> { set() }; say [33mâ��[31m∅[0mâ�¤    expecting any of:â�¤        argument listâ�¤        prefix or term…»
00:31 TimToady m: sub term:<§> { set() }; say §
00:31 camelia rakudo-moar b20535: OUTPUT«set()␤»
00:31 TimToady hmmmmm
00:32 TimToady m: say '∅'.chars
00:32 camelia rakudo-moar b20535: OUTPUT«1␤»
00:36 ventica_desktop ur thinking possibly encoding issue?
00:36 TimToady for some reason the definition of ∅ is commented out
00:36 TimToady in src/core/set_operators.pm
00:36 TimToady was, but doesn't seem so
00:37 TimToady m: say '§' ~~ /\w/
00:37 camelia rakudo-moar b20535: OUTPUT«Nil␤»
00:38 geekosaur there was a time when rakudo's parser went insane with non-ascii in the setting, wasn't there?
00:38 TimToady why would § work but not ∅?
00:38 geekosaur (20x slower or something like that)
00:38 ventica_desktop maybe camelia is a liberal arts major
00:38 TimToady which is why the setting uses \x2205 instead
00:38 geekosaur ah
00:39 rurban joined #perl6
00:39 geekosaur I was wondering since I recall ∅ being around for a while, and wondered if someone commented it while trying to track down the slowness issue and forgot to uncomment, or something
00:43 TimToady the comment seems to indicate some other issue with ACCEPTS
00:48 ventica joined #perl6
00:53 TimToady cxreg: re http://irclog.perlgeek.de/​perl6/2014-07-14#i_9023077 I would point out that such a ruling would disallow the standard < or > comparison ops, and that regex also contain the unbalanced assertions <( and )>, nor do « and » need to come in matched pairs in either regex or hypers
00:53 TimToady (now only 9 days behind on the backlog...)
00:57 TimToady ((if you guys would just quit doing stuff and talking about it, this would be a lot easier... :))
00:58 wtw joined #perl6
00:59 raydiak I *wish* I was only 9 days behind on the backlog :)
01:00 dayangkun joined #perl6
01:00 jack_rabbit joined #perl6
01:02 thou joined #perl6
01:03 raiph joined #perl6
01:05 raydiak hrm...and I'm having a heck of a time getting htmlify to run on the docs
01:07 raiph .tell timotimo https://gist.github.com/raiph/c1202f874c2477b09d1f done thru 21st, ie monday. (i think you should just do thru monday :)
01:07 yoleaux raiph: I'll pass your message to timotimo.
01:10 raiph raydiak: have you seen http://irclog.perlgeek.de/​perl6/2014-07-07#i_8981246 ?
01:11 nbrown joined #perl6
01:11 raydiak raiph: no...wonder if it's related...my errors are different
01:12 FROGGS_ joined #perl6
01:13 raiph raydiak: what are you seeing?
01:15 raydiak raiph: one gives me something quite mysterious: "Iteration past end of grapheme iterator"...the other just dies towards the end
01:15 raydiak 2 different 64-bit linux machines
01:15 woosley left #perl6
01:16 raydiak ah, I just figured the one out...apparently the process needs more than 4 gigs :P
01:16 raydiak (the one that dies towards the end)
01:16 raiph raydiak: did ya catch that jnthn did an overhaul of string handling on moar?
01:16 raydiak raiph: I've been gone for about...half a year
01:22 raydiak looks like I stopped in a couple times earlier this year, according to the log...but anyways, yeah, looks like I bettercatch up on news and such
01:25 raiph I'm really looking forward to 2014 advent calendar entries related to speed and imagining one comparing Pray speed and features as of Dec Nth 2013 with those as of Dec Nth 2014 would be perfect. :)
01:26 rindolf joined #perl6
01:26 raiph anyhow, gotta run, goodnight, great to see raydiak's bak
01:26 raydiak that is an idea...thanks raiph, g'night!
01:27 rindolf Hi all.
01:27 raydiak hi rindolf
01:27 rindolf TimToady: how are you feeling?
01:30 TimToady I'm feeling fine; just can't do too much with my eyes.
01:31 TimToady wasn't allowed to read at all the first week
01:31 TimToady they'd really prefer me not to knock my retina loose again
01:31 TimToady and I'm not allowed to change altitude
01:32 TimToady since I have a gas bubble in my eye, perfluoro propane, in fact
01:33 TimToady my last bubble was sulfur hexafloride, so I'm not sure whether that's progress...
01:34 TimToady *fluor
01:35 TimToady well, two more fluorines, anyway, but trading sulfur for 3 carbons seems kinda mundane
01:40 * TimToady wanders off to rest his eyes
01:42 colomon o/
01:45 raydiak \o
01:45 skids joined #perl6
01:47 klapperl joined #perl6
01:56 dalek doc: 1e3415f | raydiak++ | template/search_template.js:
01:56 dalek doc: Auto-highlight first result in HTML search
01:56 dalek doc: review: https://github.com/perl6/doc/commit/1e3415f1e5
01:57 figoe joined #perl6
01:57 perigrin joined #perl6
01:58 nbrown joined #perl6
01:59 perigrin left #perl6
01:59 nbrown_ joined #perl6
02:04 oJustDust joined #perl6
02:05 rindolf TimToady: bye.
02:05 rindolf raydiak: hi.
02:09 noganex joined #perl6
02:09 silug joined #perl6
02:10 cognome joined #perl6
02:15 hoverboard joined #perl6
02:20 chenryn joined #perl6
02:21 atroxaper joined #perl6
02:23 dayangkun_ joined #perl6
02:25 nbrown joined #perl6
02:30 kaare_ joined #perl6
02:31 lustlife joined #perl6
02:45 dayangkun joined #perl6
02:53 figoe joined #perl6
03:08 rurban1 joined #perl6
03:09 rurban2 joined #perl6
04:03 dalek roast/S26-WHY: 042b832 | (Rob Hoelz)++ | S26-documentation/multiline-leading.t:
04:03 dalek roast/S26-WHY: Fix up multi-line leading decl comment test
04:03 dalek roast/S26-WHY: review: https://github.com/perl6/roast/commit/042b8329cc
04:03 dalek roast/S26-WHY: 94c9408 | (Rob Hoelz)++ | S26-documentation/multiline-trailing.t:
04:03 dalek roast/S26-WHY: Start on multi-line trailing decl comment test
04:03 dalek roast/S26-WHY: review: https://github.com/perl6/roast/commit/94c9408794
04:03 dalek roast/S26-WHY: 2429965 | (Rob Hoelz)++ | S26-documentation/multiline-leading.t:
04:03 dalek roast/S26-WHY: Test nested definitions for multiline leading
04:03 dalek roast/S26-WHY: review: https://github.com/perl6/roast/commit/24299650f5
04:03 dalek roast/S26-WHY: 4fa875e | (Rob Hoelz)++ | S26-documentation/why-leading.t:
04:03 dalek roast/S26-WHY: Make sure we don't have docs when we don't expect them
04:03 dalek roast/S26-WHY: review: https://github.com/perl6/roast/commit/4fa875e075
04:03 dalek roast/S26-WHY: f6b14c5 | (Rob Hoelz)++ | S26-TODO:
04:03 dalek roast/S26-WHY: More plans
04:03 dalek roast/S26-WHY: review: https://github.com/perl6/roast/commit/f6b14c5d51
04:18 atroxaper joined #perl6
04:20 anaeem1_ joined #perl6
04:31 Alina-malina joined #perl6
04:32 chenryn joined #perl6
04:37 Alina-malina joined #perl6
05:10 bcode joined #perl6
05:12 rurban1 joined #perl6
05:14 lue joined #perl6
05:27 gfldex joined #perl6
05:36 Alina-malina joined #perl6
05:37 rurban1 joined #perl6
05:38 Alina-malina joined #perl6
05:39 dayangkun joined #perl6
05:58 atroxaper joined #perl6
06:03 kaleem joined #perl6
06:08 dalek perl6-bench: bc258b6 | (Geoffrey Broadwell)++ | / (3 files):
06:08 dalek perl6-bench: Add --skip-incomplete option for analyze
06:08 dalek perl6-bench:
06:08 dalek perl6-bench: Skips test results that have data for some compilers and not others
06:08 dalek perl6-bench: when computing summary scores; this allows comparison of compilers
06:08 dalek perl6-bench: that didn't all complete every test successfully.
06:08 dalek perl6-bench:
06:08 dalek perl6-bench: This option is defaulted on by `bench compare` and `bench history`.
06:08 dalek perl6-bench: review: https://github.com/japhb/pe​rl6-bench/commit/bc258b6530
06:08 rurban1 joined #perl6
06:13 denis_boyun_ joined #perl6
06:18 timotimo ah, that'll also make nqp/rakudo comparisons interesting
06:18 yoleaux 01:07Z <raiph> timotimo: https://gist.github.com/raiph/c1202f874c2477b09d1f done thru 21st, ie monday. (i think you should just do thru monday :)
06:28 sergot carlin: I'm surprised then :)
06:28 sergot hi #perl6 o/
06:44 zakharyas joined #perl6
06:53 lizmat joined #perl6
06:57 [Sno] joined #perl6
07:00 darutoko joined #perl6
07:02 woolfy joined #perl6
07:06 japhb timotimo: Yup, that's part of the reasoning for that change.  When I ran a big all-of-2014 bench run, and then found NQP was -- for every entry in the historical comparison, I decided to fix that.  :-)
07:06 japhb timotimo: Are you still doing the perl6 summaries?
07:09 rurban1 joined #perl6
07:13 Ven joined #perl6
07:18 Akagi201 joined #perl6
07:26 Alina-malina joined #perl6
07:39 atroxaper joined #perl6
07:46 woolfy joined #perl6
07:46 SamuraiJack__ joined #perl6
07:47 airen joined #perl6
08:02 Ven m: sub f(*%f) { say %f.perl }; f('a' => 1)
08:02 camelia rakudo-moar 1b2502: OUTPUT«Too many positional parameters passed; got 1 but expected 0␤  in sub f at /tmp/pK8uk2C46n:1␤  in block  at /tmp/pK8uk2C46n:1␤␤»
08:02 Ven m: sub f(%f) { say %f.perl }; f('a' => 1)
08:02 camelia rakudo-moar 1b2502: OUTPUT«"a" => 1␤»
08:06 Alina-malina joined #perl6
08:06 telex joined #perl6
08:10 rurban1 joined #perl6
08:12 rurban2 joined #perl6
08:15 dmol joined #perl6
08:24 xinming_ joined #perl6
08:24 bjz_ joined #perl6
08:41 chenryn joined #perl6
08:45 fhelmberger joined #perl6
08:48 dakkar joined #perl6
08:48 atroxaper joined #perl6
08:56 masak good antenoon, #perl6
08:58 Alina-malina joined #perl6
08:58 dayangkun joined #perl6
08:58 moritz good masaknoon, ante
08:59 masak and a good masaknoon it is!
09:00 lizmat joined #perl6
09:00 masak lizmat: is OSCON still going on? how is/was it?
09:01 nwc10 masak: internet says that today is the last day of OSCON. Last thing on schedule is at 1:15pm
09:01 nwc10 http://www.oscon.com/oscon2014​/public/schedule/detail/35859
09:02 masak ah, ok.
09:02 masak nwc10++ # now I feel slightly bad for having someone else check for me :)
09:02 nwc10 and 24 hours later Google docs still doesn't support ODF. I thought that web businesses ran at internet speed :-)
09:02 nwc10 masak: I knew the answer already. I just dug up the link for the exact time
09:03 nwc10 it seems to be Sunday -> Thursday for the first time this year
09:03 masak I see.
09:04 masak TimToady: I don't think I had seen ((this solution :)) to the https://xkcd.com/541/ problem before -- logically, I don't like it, but esthetically I do. :)
09:07 psch joined #perl6
09:10 ventica or {maybe this :)}
09:10 masak "interesting" :)
09:12 rurban1 joined #perl6
09:22 FROGGS_ .tell lizmat in some cases you can quote or escape paths, it does not work for all build systems but we should try
09:22 yoleaux 22 Jul 2014 23:26Z <japhb> FROGGS_: Since you used a new perl6-bench test to show the advantage of your charrange changes, can you commit and push said new test please?  :-)
09:22 yoleaux FROGGS_: I'll pass your message to lizmat.
09:30 Ven masak: (I usually do this :).)
09:30 Ven (and it's funny because he actually did a ted talk)
09:31 masak oh, so he did!
09:31 masak but about the what-if posts, it seems.
09:31 Ven Yeah, about "what if" however.
09:34 FROGGS_ .tel japhb here is the test, I'd committed it if I could make a P5 version: http://paste.scsys.co.uk/409315
09:34 FROGGS_ .tell japhb here is the test, I'd committed it if I could make a P5 version: http://paste.scsys.co.uk/409315
09:34 yoleaux FROGGS_: I'll pass your message to japhb.
09:41 timotimo japhb: i'm considering skipping it this week and making a rakudo star release post instead
09:53 eternaleye joined #perl6
10:02 sergot .seen ajs
10:02 yoleaux I haven't seen ajs around.
10:08 treehug88 joined #perl6
10:10 Alina-malina joined #perl6
10:13 rurban1 joined #perl6
10:17 kivutar joined #perl6
10:17 atroxaper joined #perl6
10:19 lustlife joined #perl6
10:19 skids joined #perl6
10:20 bjz_ joined #perl6
10:23 Mouq joined #perl6
10:23 Mouq Pls break: https://gist.github.com/Mouq/13b4d684c58b4c6add9b
10:24 Mouq i.e. try to find cases where it has incorrect behaviour
10:26 Mouq Actually... I wonder why :delete doesn't delete the whole array, the way I have that coded...
10:26 Mouq Oh, nvm
10:28 Mouq Ahh, it looks like :kv, :p, :k, etc. are broken. I didn't think to test them
10:31 Mouq Though... I'm not sure exactly what exactly the behaviour of :kv etc sould be here...
10:31 Mouq s/exactly//
10:39 tadzik hmm, is there anyone regularly using Panda on windows?
10:40 tadzik I'm looking at https://github.com/tadzik/panda/pull/91, and it surprises me that it only came up recently
10:41 FROGGS_ that patch looks very valid
10:41 tadzik huh, I just merged a pandabrew PR, and until the last moment I thought it's for rakudobrew
10:41 FROGGS_ I think I had that also that it had problems with deps or build order
10:41 masak I tried to install Rakudo on Windows to duplicate this problem, but I gave up because I remembered half way that I hate installing things on Windows.
10:41 tadzik now I wonder: "people use pandabrew? It works? :o"
10:42 masak I also approve of carlin++'s fix, and I think it should be applied.
10:43 masak but I would prefer if there were a long-term solution for this in IO::Path itself.
10:43 tadzik agreed
10:43 tadzik on both parts
10:43 colomon +1
10:43 masak IO::Path should be dealing with platform-dependent path separators.
10:43 FROGGS_ well, at least some stuff is normalized to forward slashes
10:43 tadzik but, does it need to be <$slash>, not just $slash?
10:43 masak tadzik: <$slash> is only if it's a string.
10:43 masak this is a regex.
10:43 tadzik aha
10:44 tadzik indeed
10:44 dalek panda: ff45e43 | Carlin++ | lib/Panda/Builder.pm:
10:44 dalek panda: Module paths contain backslashes on Windows
10:44 dalek panda:
10:44 dalek panda: This should allow panda to build/install on Windows.
10:44 dalek panda: Currently the required replacements are not done on Windows paths.
10:44 dalek panda: This causes panda to not work out the dependencies properly and thus not compile
10:44 dalek panda: the modules in the right order.
10:44 dalek panda: review: https://github.com/tadzik/panda/commit/ff45e43212
10:44 dalek panda: 3f8d8b3 | tadzik++ | lib/Panda/Builder.pm:
10:44 dalek panda: Merge pull request #91 from carbin/master
10:44 dalek panda:
10:44 dalek panda: Module paths contain backslashes on Windows
10:44 dalek panda: review: https://github.com/tadzik/panda/commit/3f8d8b3f28
10:44 FROGGS_ <$slash> is like evaling a string to a regex
10:45 FROGGS_ m: my $reg = /\w+/; say "abc" ~~ /$reg/
10:45 camelia rakudo-moar 1b2502: OUTPUT«「abc」␤␤»
10:49 brrt joined #perl6
10:52 atroxaper joined #perl6
11:00 chenryn joined #perl6
11:14 rurban1 joined #perl6
11:17 carlin re: panda on Windows, I was trying to start using rakudo at $dayjob when I came across that problem and was also surprised that it didn't work on Windows
11:17 carlin not all modules will break from being compiled in the wrong order though, so it's possible that's why it wasn't caught earlier
11:17 nwc10 hackathon reminder: http://act.useperl.at/apw2014/talks
11:18 nwc10 sign up! sign up! Send no money now. http://act.useperl.at/apw2014/stats
11:21 carlin ( and the error messages caused by wrong-order compiliation made me think it was a moar problem )
11:23 atroxaper_ joined #perl6
11:27 Mouq Hmm... Moar dies with Stage parse      : moar(50920,0x7fff77cde310) malloc: *** error for object 0x7fd4cb72cbe0: pointer being freed was not allocated when I use this patch on Rakudo: https://gist.github.com/Mouq/4d320515a1a718a75b22
11:27 salv0 joined #perl6
11:27 Mouq It's late now for me, but if no one picks it up, I try to look at it tomorrow
11:28 Mouq \o
11:30 carlin if you have Foo.pm6 that `use Bar.pm6` and try to compile it, rakudo whinges that you have to compile its deps first
11:30 carlin bit if it's `use Bar::Baz` moar will happily compile it without Bar/Baz.pm6 being compile first
11:35 pochi joined #perl6
11:41 rurban1 joined #perl6
11:42 Ven joined #perl6
11:43 Ven masak: we can't return a string from macro atm, can we ?
11:43 Alina-malina joined #perl6
11:45 Ven ( I'm reallyy just trying to do something like https://github.com/manastech/crystal/commi​t/0a24914e5ecee288abce21e8d3462b6dfa15f7cb​#diff-9fc1ac5e1582956c996fd04dde1f3049R73 )
11:51 prevost joined #perl6
11:54 masak Ven: can't return a string, and -- I keep repeating until, hopefully, I believe it myself -- not my department :)
11:54 masak I'm the guy who does AST macros, thank you very much.
11:55 Ven well, how can I build AST nodes ?
11:56 masak quasi {}
11:56 masak I know that's not a complete answer, though. I'd be happy to map out a nice API with someone for building AST nodes.
11:56 masak kind of a combinator library.
12:00 yakudza joined #perl6
12:07 Ven masak: well, how I can do a macro like `has-fields('a', 'b', 'c')` that does `has $.a; has $.b; has $.c;` ?
12:08 masak well, think about what that means.
12:08 masak m: has $.a
12:08 camelia rakudo-moar 1b2502: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/suOQAaG0Qtâ�¤You cannot declare attribute '$.a' here; maybe you'd like a class or a role?â�¤at /tmp/suOQAaG0Qt:1â�¤------> [32mhas $.a[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        scoped…»
12:08 betterworld can't you somehow make calls like QAST::Op.new() accessible from macros, in order to build AST nodes?
12:08 masak betterworld: I don't want to tie macros to QAST, necessarily.
12:09 masak betterworld: QAST is an implementation detail of the Rakudo line of compilers.
12:09 masak Ven: anyway, as I was beginning to say,
12:10 masak Ven: the quasi environment would need to suppress certain "environmental checks" such as the one you see in that eval.
12:10 masak Ven: that's all planned for Q4. it's a rather big job, but I think it's possible.
12:10 masak er, D4*
12:11 masak see http://news.perlfoundation.org/2011/0​9/hague-grant-application-implem.html
12:14 [Coke] are we using american english for the specs? (just saw some stuff get changed as "typos", but it could have just been british)(
12:14 masak we are definitely using American English for the spec.
12:15 moritz Britsh Typous
12:16 masak ==SOURRY==
12:18 carlin ==Very sorry, old chap==
12:18 carlin Jolly good show, but we have some potential difficulties on line 1
12:18 Ven masak: you got D1 done, right ?
12:19 jnap joined #perl6
12:20 moritz masak: did you mean, === EXCUSE ME === ?
12:21 masak Ven: D1 and D2.
12:21 Ven masak: oh, splicing is like {{{ }}} ?
12:21 masak aye.
12:22 masak the syntax everyone loves to hate :>
12:22 hoelzro morning #perl6
12:22 masak hoelzro! \o/
12:24 hoelzro ahoy masak
12:24 Ven masak: I don't love to hate it, I just can't understand what it does / is supposed to do

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

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs