Summer of Perl | 6.2.12 released! | paste: sial.org/pbot/perl6 | pugs.blogs.com | pugscode.org | pugs.kwiki.org | www.treehugger.com/files/th_images/paradigm.jpg
Set by nothingmuch on 12 July 2006.
clkao [particle]: why is todo a hash? 00:00
[particle] clkao: it's for custom attributes, todo is one of them 00:04
todo => 'reason', tags => 'op regex v6', ...
this style is used in parrot's test suite 00:05
so i just continued that here
clkao uhm. not in pugs Test.pm 00:07
:$todo
[particle] heh, duh.
ludan for $i (1..4) { say $i; } <= why does it not work on pugs? 00:16
mugwump ?eval for 1..4 -> $i { say $i }
00:16 evalbot_11895 is now known as evalbot_11910
evalbot_11910 OUTPUT[1 2 3 4 ] undef 00:16
ludan argh 00:17
is no longer possible to write a for like before?
mugwump it's a major language revision, expect everything except the general feel to change 00:18
well, that's an exaggeration
?eval for (1..4) { say $_ } 00:19
evalbot_11910 OUTPUT[1 2 3 4 ] undef
mugwump ?eval for (1..4) { say $^i }
evalbot_11910 OUTPUT[1 2 3 4 ] undef
rodi mugwump: could you remind me what $^i means?
everything made sense until that one :-D 00:20
mugwump it's an implicit argument
for (1..4) { say "$^i, $^j" }
?eval for (1..4) { say "$^i, $^j :) " }
evalbot_11910 OUTPUT[1, 2 :) 3, 4 :) ] undef
rodi that is a beautiful thing.
mugwump isn't it just
I sure don't miss the old for form 00:21
rodi I'm only on Synopse 3 :)
I HATED the old for,
and I'd been using some version of it since 1992.
mugwump ?eval loop($i=0;$i<4;$i++) { say $i }
evalbot_11910 Error: Undeclared variable: "$i"
mugwump ?eval loop(my$i=0;$i<4;$i++) { say $i } 00:22
ludan coooool
evalbot_11910 OUTPUT[0 1 2 3 ] undef
rodi salivates
ludan this implicit value is very cool $^
rodi is pavlov's lambdacamel...
?eval sub bar {<1 2 3 4>}; for bar() -> $i { say $i } 00:24
evalbot_11910 OUTPUT[1 2 3 4 ] undef
ludan it's fantastic 00:25
but not so easy to read at the first impact :/
mugwump perl has always been a bit like that... 00:26
ludan yep :P
mugwump but the funny thing is, which parts you find easy to read and which seem natural depend on your background
to a ruby programmer, the pointy sub notation ( -> $x { ... }) would seem very natural
ludan for me it's the first time 00:27
but in your opinion, it's better to have (1..4) -> $x than $x ( 1 .. 4 ) ?
00:29 lollan joined
[particle] no spaces allowed between subname and parens in p6, rats. 00:40
one bug squashed.
rodi [particle]: aren't you allowed to do foo\ .() # or is this not what you were talking about? 00:44
[particle] you are, i had foo () 00:46
rodi right-o
[particle] it was two spaces, so not really worth doing \.
00:47 toku_away is now known as tokuhirom 00:52 hikozaemon joined
meppl gute nacht 01:02
sahadev2 hello, i am getting a bunch of errors like: src/DrIFT/YAML.hs:43:50: Not in scope: data constructor `ENil' 01:11
compiling rev 11910 01:12
any ideas?
01:12 Khisanth joined 01:20 christo joined 01:29 justatheory joined 01:35 dolmans joined 01:37 christo left
svnbot6 r11911 | clkao++ | hack for support /$pattern/ interpolation 01:56
01:58 weinig|away is now known as weinig 02:14 weinig is now known as weinig|sleep
dolmans for (@foo, @bar) {} just traverse the two array object, how to flatten them in place? 02:44
02:44 leoncamel_ joined
clkao [particle]: still around? 02:48
revdiablo dolmans: for (*@foo, *@bar) {}
02:50 leoncamel_ left
dolmans revdiablo: cool, thanks. 02:51
revdiablo Steamroller operator =) 02:53
Or maybe that's **, I forget
But *@foo flattens
dolmans i tested *, it works fine. 02:54
svnbot6 r11912 | clkao++ | p6regex.t now reads from the test spec file.
r11911 | clkao++ | hack for support /$pattern/ interpolation
r11913 | clkao++ | EXECUTABLE_NAME or PROGRAM_NAME? Both for now. 02:56
02:57 vel joined
clkao alright, done with pugs hacking today 02:57
svnbot6 r11914 | clkao++ | Make v6 compile (and segfault while running) p6regex.t 02:59
r11914 | clkao++ | * Implement Perl6::Internals::open.
r11914 | clkao++ | * Support prefix = for reading a line from file.
03:23 amnesiac joined
svnbot6 r11915 | clkao++ | Fix binding for other mg. 03:30
clkao fglock: we can now run regex_tests 03:34
svnbot6 r11916 | clkao++ | Use $/ directly. Passing 193 tests of 485.
clkao oops, it's failing 193 tests out of 485 03:35
03:47 lisppaste3 joined
TreyHarris ?eval my $x = 0 but true; say "yes" if $x 03:59
03:59 evalbot_11910 is now known as evalbot_11916
evalbot_11916 Error: No such sub: "&true" 03:59
TreyHarris what did "0 but true" become? I remember it changed, but I don't remember to what
buu ?eval my $x 0 but TRUE 04:19
evalbot_11916 Error: unexpected "0" expecting comment, "?", "!", trait, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
buu er
?eval my $x = 0 but TRUE
evalbot_11916 Error: No such sub: "&TRUE"
04:24 stevan joined 04:44 jlhamilton joined 04:51 shachaf joined 04:55 lollan joined, lollan_ joined 04:57 Khisanth joined 05:01 jlhamilton left
wolverian ?eval True 05:05
evalbot_11916 Bool::True
wolverian ?eval my $x = 0 but True
evalbot_11916 Error: Cannot cast from VBool True to Pugs.AST.Internals.VCode (VCode)
05:14 [particle] joined 05:38 kanru joined 05:55 rgs joined 05:58 Aankhen`` joined 06:07 buetow joined
gaal sahadev2: could you nopaste more context to the error? looks like you're trying to compile a file you shouldn't be compiling. 06:16
dolmans what dose the perl6/pugs ffi look like? especially call c libs. 06:40
06:49 pwnny joined
pwnny z6.invisionfree.com/Best_Damn_Sport...hp?act=idx 06:49
06:51 pwnny left
gaal in vim, how do I get the unicode code point for the character under the cursor? 'g8' gives utf8 value, which is not quite what I want. 07:03
07:04 iblechbot joined
gaal ah, 'ga' works. 07:05
In S02, there's an example: my Dog $spot; say $spot; # # "Dog" 07:16
Shouldn't that better print "::Dog" ?
07:29 lambdabot joined
gaal hmm, maybe not, because of the binding property of ::... 07:34
07:41 baest joined
gaal should the paragraph in S02 beginning "All prefix sigil operators accept one positional argument, evaluated in scalar context as a rvalue." be raised to its own bullet item? 07:43
08:05 iblechbot joined
clkao is it open($fh, ...) or $fh = open(...) ? 08:12
dolmans the latter i saw in examples, but don't know the prior works or not. 08:17
clkao it's everywhere in t/ 08:20
oh no nevermind
i am crazy. it's in the .tc file. of course.
dolmans yes, it's in t/builtin/io/... 08:21
gaal sure, the latter, and it's a first-class object. 08:23
clkao *nod* 08:24
gaal: v6 already implements that
if you haven't noticed
08:27 marmic joined, ruoso joined
dolmans the haskell fans scared me so much, they talked about "programming is essential mathematics". Buddha save me, i am always in the impression that programming is literature and art. 08:29
azuroth did they say that isn't what maths is? 08:30
kattana_ Different people see beauty in different fields.
TreyHarris clkao: did you see my question earlier? how is "0 but true" written nowadays?
dolmans kattana_: indeed, oh, somebody says math is also art, so i laughed. 08:31
kattana_ fractals are math and art.
clkao ?eval my $x = 0 but True; say "yes" if $x 08:32
evalbot_11916 Error: Cannot cast from VBool True to Pugs.AST.Internals.VCode (VCode)
clkao bug
dolmans and da vinchi's paint is math too.:) 08:33
TreyHarris clkao: ok, good
kattana_ Lots of painters actualy used math in their paintings, precise color and contrast, only without knowing it. 08:34
But is it beautiful because it is mathamaticaly precise, or fits the math because it is beautiful? 08:35
dolmans this is a philosophy question 08:36
azuroth your mother is a philosophy question
dolmans azuroth: what do you mean?
kattana_ Everything is if you twist the question the right way, thats the beauty of philosophy ;)
azuroth umm... I mean that the concept of life and birth is a philosophical one 08:37
dolmans that's true.
but otherwise is not.
rgs programming little algorithms might be maths, but programming large projects is definitively literature, and often not so good literature. 08:38
crime stories mostly (when you try to discover the murdere^Wbug)
clkao rgs++ 08:40
with perl6 hopefully it's going to poetry
azuroth crime stories written in pig latin...
clkao to be poetry
dolmans i hope too.
azuroth hmm. I have a whole weekend to commoditise - what to do! 08:42
rgs go to the moose zoo ! 08:44
clkao make 100 tests pass for v6
azuroth one hundred?? want to show me where to start? :-p 08:45
clkao i just implemented IO a few hours ago... making stuff under builtin/io pass might be a good start 08:46
er i meant i implemented open
thunking the methods etc
azuroth humm, okay. I'll need to find the v6 repo
clkao it's pugs/perl5/Pugs-Compiler-Perl6 08:47
azuroth oh. sweet
clkao ./make_build_perl5; perl util/src_to_blib.pl
then you can start util/prove6 t/builtin/io
clkao goes to bed 08:48
azuroth I'll see how far I can get...
pasteling "cmarcelo" at 200.232.237.151 pasted "Implement IArray and IHash using HsJudy. Tested against r11916." (535 lines, 21.9K) at sial.org/pbot/18775 08:51
08:53 nothingmuch joined
svnbot6 r11917 | cmarcelo++ | * Just hides some debug messages. 08:53
cmarcelo audreyt: "pugs -e 'my @a; @a[999999999] = 2'" works here with this patch.. I tested things in t/0* t/builtins t/data_types and all seem to be broken (btw, they were VERY helpful in debugging). i'll let two smokes (trunk and my local version) running here while I sleep to see if anything is damaged. 08:55
08:57 elmex joined
cmarcelo audreyt: a strange thing is [email@hidden.address] "pop @a" and others are being trapped somewhere before array_pop or array_fetchSize being called, i don't know where =P.. 08:58
oh, just in case: "./pugs -e 'my @a; @a[999999999] = 2; say @a[999999999];'" works too =) 08:59
also, what's that comment in IArray's array_storeElem and array_fetchElem saying "wraparound"? 09:00
oh my: s/all seem to be broken/all seem to be broken only where original trunk were broken/ 09:02
azuroth aha! that makes more sense :D 09:04
cmarcelo =D
09:04 revision17_ joined
nothingmuch audreyt: around? 09:09
gaal: around? 09:14
09:15 sqd joined
sqd hi 09:15
nothingmuch hi
sqd could someone help me please with a little conversion`? 09:16
nothingmuch is willing to try
sqd can I paste 6 lines?
nothingmuch use the link in /topic
pasteling "sqd" at 129.16.101.79 pasted "script 1 (old working version)" (6 lines, 377B) at sial.org/pbot/18776 09:18
sqd this perl script is converting java script from a website
and now that website was change and I can't manage with converting perl script 09:19
changed*
nothingmuch FYI, this channel is for perl 6 discussion, not perl help 09:20
what's dk42?
sqd it's a js function
nothingmuch that you're parsing??
sqd I'm not familiar with that function 09:22
nothingmuch okay, nevermind
so you're trying to not match the additional ' + ' ?
sqd script 1 was changed and is working now
and I need to convert wcript 2
nothingmuch the perl is just matching 09:23
sqd dk42 was replaced by write in js
js: dk42( unescape( 'some key' ) ) perl: ~ /dk42\( unescape\( '([^']+)/; 09:24
js: + write( unescape( 'some key' ) ) perl: this is the question ;)
nothingmuch oh
s/\bdk42\(/write\(/gs; 09:25
that should do the trick
sqd $line=s/\bdk42\(/write\(/gs;
right?
nothingmuch yep 09:26
but be careful
that'll also replace dk42( in strings
sqd there is no dk42 in new js
nothingmuch i menat in the old js
sqd there is "write" instead of "dk42
nothingmuch to do a 100% safe find/replace you need to parse the js with a grammar
and replace only the function call nodes 09:27
sqd old is working, I need to fix the new one
azuroth does < and > work in perl regex like in vim..? ^_^;
nothingmuch oh
so do you want to replace all the write() to dk42() ?
sqd no no
I have to get the key from js using perl
in old version of js it was done like shown in the file 09:28
gaal nothingmuch: moose
nothingmuch gaal: i want someone to talk to about composed interfaces for Moose
or actually Class::MOP
sqd and now admin changed the js replacing function dk42 with "write"
nothingmuch sqd: the file doesn't do any ocnversion
the nopaste, anyway
just matching
sqd and I don't know how to fix that in perl
gaal nothingmuch: shoot 09:29
azuroth it captures the key in ()s?
nothingmuch gaal: #moose?
gaal you got it
sqd nothingmuch what makes this line: $line =~ m#\+ write\( unescape\( '([^']+)#) with this js: dk42( unescape( 'some key' ) ) ? 09:30
nothingmuch sqd: i don't understand the question
sqd to get "a key" form the website I have to use perl script 09:31
in the js on the website was a piece: dk42( unescape( 'some key' ) ) 09:32
respective perl script to get the key was: ($line =~ m#dk42[^']*'([^']*)'#)
now the js on the website was changed to: + write( unescape( 'some key' ) ) +
and respective working perl script to get the key is: ($line =~ m#\+ write\( unescape\( '([^']+)#) 09:33
my question is:
I have another perl script to get that key
cmarcelo I did a "./make_build_perl5; perl util/src_to_blib.pl", make install and then tried perl t/01-sanity/01-tap.t and got a lot of errors like: String found where operator expected at t/01-sanity/01-tap.t line 6, near "say '1..10'" (Do you need to predeclare say?)
sqd for old js version is was: $line=~ /dk42\( unescape\( '([^']+)/;
cmarcelo anyone knows what is causing this? 09:34
sqd how to change the second perl script to work for the new js.
azuroth just change the dk42 to write? 09:35
sqd no
it doesn't work
working example is shown above 09:36
but I don't know how to fix script 2
($line =~ m#dk42[^']*'([^']*)'#) <--for dk42 09:37
($line =~ m#\+ write\( unescape\( '([^']+)#) <--- for "write"
this was script 1
how to do the same in script 2?
both scripts are here: sial.org/pbot/18776 09:38
lambdabot Title: Paste #18776 from "sqd" at 129.16.101.79
azuroth I don't see why the first wouldn't still work.. 09:39
sqd the first is working
but forsome reason I have to use script 2
script 1 was succesfully converted, my task is to do the same with script 2 09:41
azuroth hmm. well the first regex matches the second write( etc ) for me.. 09:42
sqd hmm 09:44
what makes m# in ($line =~ m#\+ ? 09:51
azuroth m# ... # is just like m/ ... / or m( ... ) 09:52
sqd # replaces whatever? 09:53
09:55 kane-xs joined
azuroth # is just hte delimiter 09:55
only works for regexes and qw and stuff, I think
sqd azuroth, could you please explain that line: ($line =~ m#\+ write\( unescape\( '([^']+)#) 09:56
azuroth how much of it do you already understand? :-)
sqd a little
I'm newbie
let'ōæ½xA8s say almost nothing :D
I'm not familiar with perl 09:57
azuroth well variables are preceded by $.. so $line is a variable. =~ is the regex operator - do you know about regular expressions? 09:58
StudiesInWords try #perlhelp
sqd I don't know, but about regex
I don't know about regex
(sorry)
I want to undestand that piece: m#\+ write\( unescape\( '([^']+)#) 09:59
what does it make with "line" variable
StudiesInWords there is only one person at #perlhelp 10:00
nothingmuch #perl
azuroth it just takes $line's value (say the string "foo bar + write( unescape" ) 10:01
StudiesInWords efnet/#perlhelp (right?)
sqd right 10:02
10:04 rodi joined
sqd thanks azuroth 10:05
azuroth that's alright
dolmans just takes $line's value (say the string "foo bar + write( unescape" ) 10:06
<StudiesInWords> efnet/#perlhelp (right?)
oops,sorry.
sqd that I already know dolmans :)
dolmans :-P, that's true. 10:07
sqd azuroth do you have a minute to help me with a little problem? 10:10
azuroth sure
sqd sial.org/pbot/18776
lambdabot Title: Paste #18776 from "sqd" at 129.16.101.79
sqd I need to fix script 2 10:11
azuroth hmm, I still can't see why script 2 won't work as it is
sqd script 1 was fixed and is working, i have to do the same with script2
azuroth it should just be the same as script 1 10:12
sqd azuroth script 2 doesn't exist
azuroth well script 1 should do everything that script 2 has to 10:13
sqd that's the problem
azuroth right
script 1 was already fixed
the problem is with script2
in script 1 is a function dk42, in script 2 is write 10:16
azuroth hmm, okay... how is it a problem with script 2 if script 2 doesn't exist? :S 10:17
10:17 chris2 joined
sqd the problem is to write perl script 2 10:18
(exactly one line)
rodi In S03, the word "brainos" - does it just mean "silly mistakes" ?
azuroth like typos, yeah
except brainographical errors
rodi I have learned a lot of new syntax reading these things, just wanted to make sure I'm not missing some obscure linguistics term ;) 10:19
sqd azuroth could yopu please look at computers.ovh.org/problem1.txt ? 10:21
azuroth the regex should all be OK 10:23
sqd all are OK
all are/were(before js changes, script 1) working 10:24
#example in that file is in js 10:26
azuroth hmm, okay 10:27
I really don't know :x 10:28
sqd ok, thanks for help anyway
10:48 theorbtwo joined 11:21 marmic joined 11:37 ivas joined 11:40 SM_ax joined 11:42 ludan joined 11:46 kane__ joined 12:01 Limbic_Region joined 12:15 dolmans joined
gaal in s02, this example for adverbial modification isn't clear to me: 12:38
for =$handle :prompt('$ ') { say $_ + 1 }
why is it the read that's being modified?
is auxilary:<prompt> simply { $^v ~ $_ } ? 12:40
12:56 ludan joined
cmarcelo audreyt: re: patch, after smokes I found that t/operators/assign.t has a lot of tests broken with my version, I'm checking on that. 12:59
13:01 scook0 joined 13:03 mako132 is now known as maco132 13:11 sahadev2_ joined 13:13 mauke_ joined
kolibrie I would like to try out v6.pm, and tried clkao's instructions at use.perl.org/~clkao/journal/30407 13:15
lambdabot Title: Journal of clkao (1687)
kolibrie unfortunately, make build_perl5 seems to not find Data::Bind and other modules under pugs/perl5
and util/prove6 t/01-sanity says: 13:16
bash: util/prove6: /bin/perl: bad interpreter: No such file or directory
do I need to set some environment variable or path so that the modules under perl5 are found?
cmarcelo kolibrie: about prove6, you can change the first line to match your perl location... i.e., /usr/bin/perl 13:17
first line of util/prove6 script
kolibrie cmarcelo: hmm, well, now that appears to run, but the 01-sanity says...dubious 13:18
t/01-sanity...pugs: t/01-sanity: openFile: inappropriate type (is a directory)
cmarcelo ops
try with t/01-sanity/*.t 13:19
kolibrie cmarcelo: yes, that seems to work nicely 13:20
thank you
cmarcelo tests passed?
kolibrie cmarcelo: yes, except for two which were skipped
but that is the haskell backend
cmarcelo hmm
13:20 justatheory joined
cmarcelo util/prove6 --perl t/01-sanity/*.t 13:21
kolibrie that gets me lots of errors
cmarcelo something like this: "String found where operator expected at t/01-sanity/01-tap.t line 6"
?
kolibrie yes, like that 13:22
String found where operator expected at t/01-sanity/01-tap.t line 6, near "say '1..10'" 13:23
cmarcelo kolibrie: welcome to my world =o)... i'm stuck at this point too...
kolibrie I think that not everything from pugs/perl5 made it into the blib 13:24
or something
fglock: clkao: can you shed any insights into your v6.pm setup?
13:27 penk joined 13:28 nothingmuch joined, mauke_ is now known as mauke
cmarcelo kolibrie: i'm going to sleep... report in the channel if you find how to solve it before rescue come. tks & 13:30
fglock kolibrie: you can do a first install from cpan, and then install the newer perl5/ modules to get the latest devel version 13:35
13:39 vel joined, justatheory joined 13:43 iblechbot joined
kolibrie fglock: so you generally install, rather than just running from the pugs tree? 13:44
fglock kolibrie: the cpan is known to work, so it's easier to start with 13:45
cpan version
kolibrie thought that since he already had the pugs tree, that would be easier 13:46
fglock please install Cache::Cache also
kolibrie that makes it faster, I presume? 13:47
fglock yes, it is used to cache rules as perl5 code 13:49
kolibrie sounds nice
[particle] gah. cpan v6 fails make test on windows
it assumes the shell does globbing for you 13:50
this should be an easy fix
by putting double quotes around "t/*.t" etc
i'll post a small patch 13:54
fglock [particle]: thanks! 13:55
14:01 hexmode joined 14:07 Anil joined 14:15 robkinyon joined 14:18 robkinyon joined 14:21 chris2 joined
svnbot6 r11918 | fglock++ | pX/Pugs-Compiler-Rule - /{ return undef }/ actually returns an undef value 14:22
14:22 weinig|sleep is now known as weinig 14:27 lolage joined 14:36 tokuhirom is now known as toku_away
svnbot6 r11919 | fglock++ | pX/Pugs-Compiler-Perl6 - '--compile-only' works 14:36
r11918 | fglock++ | pX/Pugs-Compiler-Rule - /{ return undef }/ actually returns an undef value
14:38 prefiks joined, ruz joined 14:46 weinig is now known as weinig|away
[particle] okay, my changes seem to work 14:49
boy, it takes a while for these tests to run, though 14:50
fglock [particle]: yes, i hope the pX rewrite will make it faster 14:52
[particle] fglock: are there expected test failures? 14:53
14:54 amnesiac_ joined 14:55 amnesiac_ is now known as amnesiac
fglock the numbered tests should pass - 01-sanity, 02-test-pm; the pugs test suite may fail some tests 14:55
pasteling "[particle]" at 144.81.84.180 pasted "modified v6-alpha Makefile.PL for cross-platform friendliness" (59 lines, 1.6K) at sial.org/pbot/18778 14:56
[particle] okay, that's good then.
i pasted it here, it was easier to do than diff
note 'use File::Spec' and the changes in makemaker_args() and clean_files() 14:57
All tests successful (12 subtests UNEXPECTEDLY SUCCEEDED), 2 subtests skipped. 14:58
yay!
fglock [particle]: commit Makefile.PL to svn?
[particle] oh, v6-alpha-0.11 is in the pugs repo? 14:59
fglock perl5/Pugs-Compiler-Perl6/ 15:00
the Makefile.PL in svn may be different from cpan 15:01
[particle] close enough 15:02
fixed a bug in it, too 15:03
hence the 12 unexpected success
fglock :) 15:04
svnbot6 r11920 | particle++ | [Pugs-Compiler-Perl6]: modified Makefile.PL for cross-platform friendliness 15:06
r11921 | fglock++ | pX/Pugs-Compiler-Perl6 - added a source file line counter 15:12
15:12 meppl joined 15:13 frederico joined 15:16 buetow joined
[particle] t/p6regex/p6regex....Can't locate v6.pm in @INC (@INC contains: 15:23
aargh
15:25 crem_ joined 15:28 kanru joined 15:43 cjeris joined 15:44 weinig|away is now known as weinig
mindbob whats the best way to go about playing with environment variables in perl, more specifically, accepting a given env variable and displaying its memory addy? 16:05
16:07 weinig is now known as weinig|bbl 16:08 Psyche^ joined
TimToady SageLT: {xxx} in P5 doesn't treat curlies as metacharacters unless xxx =~ /\d+(?:,\d*)?/ 16:14
16:16 iblechbot joined, Psyche^ is now known as StudiesInWords
TimToady clkao: A orz B can parse as infix:<orz> only if it is predefined. 16:17
fglock is correct, the only possible parse for an undeclared orz is A( orz(B)), and that requires orz to be defined as compatible to listop by the end of the file. 16:18
svnbot6 r11922 | fglock++ | pX/Pugs-Compiler-Perl6 - fixed source file line counter for one-liners
PerlJam mindbob: This channel is about perl6. You probably want to stay on #perl
TimToady and even that parse is impossible if A is unambiguously the end of a term. 16:19
since you can never have two terms in a row.
16:20 Corion joined
TimToady dolmans: * as flattener is deprecated--we use [,] now, but it shouldn't be necessary (@foo, @bar) in list context, which should flatten anyway (lazily) 16:20
did I miss anything? 16:21
Corion What? The steamroller operator is dead? Awww :-(
Anyway, just passing through, have fun;)
TimToady if you're on the way to the fjords, maybe you'll find the steamroller there. :)
Corion TimToady: I'm not dead yet. I still have lots of voom! 16:24
16:35 weinig|bbl is now known as weinig
TreyHarris TimToady: and "0 but True" is correct, and a Pugs bug? 16:38
?eval 0 but True
16:38 evalbot_11916 is now known as evalbot_11922
evalbot_11922 Error: Cannot cast from VBool True to Pugs.AST.Internals.VCode (VCode) 16:38
clkao oh, v6 passes 1229 tests now 16:41
[particle] clkao++ for getting p6regex harness working 16:42
if only i had v6 working, i could test it :)
fglock clkao: I'm starting to s/Regex/Token/ in pX
[particle] how will C<rule dbi_ns { ::DBI }> parse? is :: a cut marker or ns sigil? 16:44
fglock {} is a closure, so it's a sigil 16:45
oops, no
16:46 Corion_ joined
fglock the rule is ' ::DBI ', so it's a cut marker 16:46
clkao [particle]: what's stopping v6 working for you?
[particle] clkao: i fixed win32 makefile issues this morning
but prove6 can't find v6 in @INC
PerlJam clkao: Is there some roadmap for future svk changes somewhere? 16:48
16:48 weinig joined
[particle] is running nmake install again, to be sure 16:48
PerlJam fglock, [particle]: having a cut marker at the beginning of a pattern is weird. 16:50
[particle] i think '::' as a sigil is weird 16:51
clkao PerlJam: if it's not on svktodo then there's not. perhaps i wrote something on the mailing list
[particle] how does C<rule { $x }> parse? (sigil or eol?)
fglock $x interpolates to a string or to a subrule 16:53
[particle] right. so does ::foo interpolate to something?
16:54 DaGo joined
clkao TimToady: ok... so the parser has to change the way parsing when it sees infix:<orz> defined 16:55
[particle] what if i have C< grammar foo; ...; package bar; rule baz { ::foo };> ?
PerlJam ::foo should probably warn IMHO
[particle] :: is also a sigil for grammar (as per S02) 16:56
can a grammar be inside a rule?
if not, s/grammar/enum/
i don't see a clear answer 16:57
PerlJam I thought the only uses of :: as sigil were in declarations and as ::(...) for interpolation 16:59
17:00 tup joined
[particle] hrmm 17:00
so rule { ::(foo) }
still ambiguous
TimToady TreyHarris: yes, 0 but True should work.
TreyHarris ok. hopefully i'll be able to write a test soon. supposedly, the application just needs final approval in general counsel's office or somesuch... the past few weeks have been really excruciating, seeing all these things i want to contribute and not being able to... :-) 17:03
17:03 frederico joined
[particle] wonders if s/True/true/ might work 17:03
PerlJam I'm having trouble seeing why anyone would want to do rule foo { ::(...) } where :: is cut, so I'm inclined to say that it never is. 17:04
17:04 GODAttach joined, GODAttach left
[particle] istr there was a case change in pugs recently, dunno if it was completed 17:04
17:05 bernhard joined
[particle] it doesn't have to be empty in front of the :: 17:05
foo::bar(baz)
TimToady clkao: yes all function definitions mutate the language to some degree. As a kind of macro, installing an infix is required to have only lexical scope. 17:06
PerlJam we have whitespace disambiguation for that.
foo::bar (:: is never cut), foo :: bar (:: is always cut)
[particle] but rules are ws sensitive 17:07
so, i suppose you need <?wb> disambiguation
TimToady what do you mean by ws sensitive?
[particle] what's the modifier... :sigspace 17:08
TimToady the inside of a rule works the same way as the outside, linguistically speaking--we still follow longest-token rule. It's just a different language
:sigspace gives extra significance to whitespace.
clkao i wonder if/how PCR can handle this
TimToady so ::foo should be reserved for types, in case we ever want to bind a type: ::foo := (...) somehow. 17:09
[particle] clkao: the parser should handle that
TimToady basically, I think backtrack controllers should almost always have a space after them.
much like label colons
[particle] oh, label colons have required post-space? wasn't aware of that 17:10
clkao ya, i know it should. just wonder if we can already handle it
[particle] oh, *already*
well, longest token rule works for me. i should have thought of that. 17:11
PerlJam [particle]: yeah, it's hard to keep everything in your head at once though :) 17:12
[particle] and if you really want to be clear, <'::'>foo
pj: this is why i ask questions :)
i've forgotten more than i'll ever know 17:13
(to turn a phrase) 17:14
PerlJam wonders if bestpractical has an RT queue setup for svk so that he could add a ticket for clkao to produce an svk roadmap ;) 17:25
[particle] joins #svk 17:26
clkao PerlJam: for now we are using hiveminder ;) you want to join svk group ? 17:29
PerlJam clkao: what does that mean to me? 17:30
17:35 tup joined
clkao PerlJam: what's your email? 17:39
PerlJam clkao: [email@hidden.address]
17:42 frederico joined
svnbot6 r11923 | fglock++ | pX/Pugs-Compiler-Perl6 - s/Regex/Token/ in several places 17:48
17:56 hexmode joined
pasteling "[particle]" at 144.81.84.180 pasted "svk debug (#perl6 please ignore, intended for #svk)" (101 lines, 3.1K) at sial.org/pbot/18779 17:56
17:59 ajs joined 18:08 xinming_ joined 18:13 vel joined 18:14 vel joined 18:22 Eimi joined 18:23 iblechbot joined 18:30 cookys joined 18:36 crem joined
[particle] figured out why v6 wasn't found in @INC 18:42
because it wasn't in @INC!
it's in PERL5LIB
19:17 crem joined 19:21 rindolf joined
rindolf Hi all! 19:21
TimToady: here?
19:25 fglock joined
rindolf audreyt: here? 19:27
19:27 bradb joined 19:38 embedded_cm joined
rindolf www.shlomifish.org/philosophy/compu...-language/ 19:38
lambdabot Title: Toughts about the Best Introducory Language
19:40 weinig joined 19:41 weinig is now known as weinig|bbl 20:08 tup joined 20:18 sahadev2 joined
svnbot6 r11924 | fglock++ | pX/Pugs-Compiler-* - implemented 'fail' inside rules 20:21
20:34 larsen joined 20:35 fglock joined
svnbot6 r11925 | fglock++ | pX/Pugs-Compiler-Perl6 - Prelude compiles 20:42
clkao whoot 20:44
svnbot6 r11926 | Darren_Duncan++ | Updated the documentation in ext/Rosetta/lib/Rosetta/SeeAlso.pod and ext/Set-Relation/lib/Set/Relation.pm
20:59 SageLT joined
SageLT audreyt: you available for a quick question on running ghc on intel macs? 21:00
21:03 fglock_ joined
fglock_ clkao: I think it is possible to start the merge 21:05
21:10 hexmode joined
clkao fglock_: excellent 21:16
svnbot6 r11927 | fglock++ | pX/Pugs-Compiler-Perl6 - cleanup debug info 21:17
clkao do the tests pass?
fglock_ I only run a few tests - I'll try in windows now 21:19
kolibrie cmarcelo: I never got back to v6.pm, I'll try again after the weekend 21:30
clkao fglock_: also have you noticed we can run p6regex.t 21:46
21:46 TimToady joined 21:47 prefiks left
clkao fglock_: fyi, tailcall is not passing 21:51
fglock_: but we might as well merge up to trunk and fix it there. 21:53
fglock_ clkao: I'll take a look 22:01
TimToady: ping
clkao ok. meanwhile i found the merge doing something stupid. i will look at svk 22:02
ETOOMANYYAKS
fglock_ notice: YAPC::SouthAmerica is moving nov 3-5 (was nov 9-11) 22:04
we got a bigger place, but we had to change dates
clkao oh this is getting interesting. 22:05
how far is it from SFO or BOS?
and do i need a visa
TimToady fglock_: pong
clkao fglock_: ah anotehr thing is can we change the parse on the fly inserting now infix prefix etc? 22:06
fglock_ TimToady: please check if the new date is ok for you
TimToady I'll have to check with Gloria. 22:07
fglock_ clkao: yes, there is a 'recompile' method in each class, that can be used to create a new local grammar
TimToady: ok 22:08
clkao: the airport code is GRU
SageLT audreyt: Are you around? I got a quick question on ghc on intel macs. 22:12
fglock_ YAPC::SA -> en.wikipedia.org/wiki/Anhembi_Convention_Center 22:14
clkao TAM has direct flight 22:16
22:18 weinig|bbl is now known as weinig 22:19 cjeris left
svnbot6 r11928 | Sage++ | This is the script I've been using for testing. It's simple and underdocumented at this point, but it runs translate across all yaml files in a directory. Since I've been testing on a Mac that has problems with ghc compiling, there's also an extremely slow option to run ghc seperately for each test (with no comiling). 22:22
fglock_ clkao: are you in US or England? 22:30
i mean, in november
22:32 xinming_ is now known as xinming
clkao fglock_: london 22:34
i don't know, might be in US
22:40 hexmode joined
fglock_ biab & 22:44
22:45 fglock_ left 23:28 weinig is now known as weinig|bbl 23:30 justatheory joined 23:49 mdiep joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!