2009 -- the year of November. <november-wiki.org> <github.com/viklund/november> <irclog.perlgeek.de/november-wiki> <nopaste.snit.ch>
Set by moderator on 7 April 2009.
Tene blogs.gurulabs.com/stephen/2009/04/...paste.html 00:49
zarah Tene's link is also tinyurl.com/c7lqut
Tene I added filesystem persistence for pastes and a recently-pasted list on the main page 00:50
live kopipasta demo at pleasedieinafire.net:2080/ 00:55
zarah Tene's link is also tinyurl.com/d7jxly
spinclad zarah: f/w? 03:48
zarah spinclad: eh... wha'?
Tene spinclad: what is 'f/w'? 03:52
spinclad never mind, i got it: f/w is framework 03:57
i saw one reference earlier that i couldn't work out, then a later use made it clear. 03:58
Tene ah 03:59
spinclad glad to catch kopipasta up now... i caught your code pastes earlier (tasty pasta indeed), then couldn't connect when i tried later. 04:14
Tene Yeah, my connection at home isn't so great 04:23
05:35 ihrd joined 06:01 masak joined
Tene masak, ihrd: I got kopipasta to save to the filesystem (needs better error handling) and encode the pastes with HTML Entities. 06:15
masak Tene: nice.
I'd like to bring it under the reign of Web::Request and Web::Response, and maybe touch up the stylesheet a bit. 06:16
Tene There's no stylesheet right now.
masak exactly. :)
Tene What I *really* want to do is write a module that will simply serve filesystem requests
masak Tene: sure, go right ahead. 06:17
Tene so I can put the stylesheet on the fs, and just redirect there
Also, btw, HTML::Entities needs a decode()
simple to get there from the encode, but I didn't need it yet
masak mberends and I talked during the weekend about adding DTD validation to Tags. we think statically converting the DTD to a piece of Perl data structure is the way to go. 06:18
Tene that would be nice 06:21
Tags also needs to be separated out into different behavior for HTML, XHTML, etc. 06:22
masak aye. 06:25
maybe time to work on a test suite?
Tene also needs a better name 06:26
'Tags' is a horrible name. :)
masak aye.
Tene also, we need a different name for 'div'
because it conflicts with the Perl 6 'div'
masak same with 'map'.
also, we need a way to warn users who just assume that 'div' and 'map' will be named that, without thinking of the Perl 6 functions. 06:27
waitwait, isn't the Perl 6 'div' an op?
Tene not sure
masak then it's no issue, methinks. the problem with 'map' is that it's a function. 06:28
Tene I was seeing something strange with 'div' and didn't investigate further
masak ok.
Tene I'm thinking we just import a few common elements (html, head, body, title, p, ul, ol, li, table, etc) and put the less common tags in a different set of exports 06:29
so you have to ask for :extra or :all or something
the other option is that default doesn't import anything, and you have to ask for :html, :xhtml, etc. 06:30
but I prefer different modules for that
but I'm convinceable
ihrd: I read through Routes.pm a few times, but I don't really get it. Are there any example uses of Routes? 06:31
oh, t/routes 06:32
Hmm... i don't much like it... will need to think a lot more about why. 06:33
maybe I will change my mind. 06:34
ihrd Tene: hi! sorry, I can`t discuss now, mail me you sugesstions plz 06:35
Tene ihrd: I can't discuss now also
:)
masak: I am considering writing an IRC bot that will eval perl 6 and post to kopipasta 06:36
masak Tene: cool.
Tene anyway, sleep time now 06:37
goodnight all
masak Tene: we have a better option nowadays than STD.pm
let's talk about it tomorrow. :)
Tene: night.
Tene yes, I would like suggestions
I have secret plans of porting POE to Parrot
masak I'll try Perl 6 parsing a bit locally.
I need it for many different things. 06:38
Tene hm?
explain?
masak which part? the one where I need to parse Perl 6? 06:39
Tene Yes, what your goal is.
masak Tene: with regard to kopipasta, colour coding. 06:40
Tene ah
masak there's no pastebin that colour codes Perl 6 yet.
Tene That would be very nice, yes.
I like that.
I may experiment as well.
masak that was the idea with making a pastebin. :)
what's new is that Rakudo can now output a parse tree.
--output=parse
Tene you can also get a match object in rakudo itself 06:41
masak Tene: matching on a Perl 6 program?
Tene Yes.
I can't remember which syntax works right now, but the grammar is in Perl6::Grammar 06:42
masak oh!
masak must experiment
I'm amazed at how such fantastic news manages to sneak past me :)
Tene masak: that's been the case as long as rakudo has been able to match regular expressions 06:43
$text ~~ /<Perl6::Grammar::TOP>/
works
kinda
masak o_O
ok, that's... perfect.
moritz_ you can try with .parse and :actions 06:44
and supply your own actions
and write a Perl 6 compiler in Rakudo ;-)
masak I'll try that right away. 06:45
oh, but it only works for the things Rakudo can parse, of course... 06:46
so I still can't use it for user-defined operators.
oh well, I'll use STD.pm for that.
Tene masak: pmichaud suggested those might be coming RSN
masak I know.
I just plan to be first. :) 06:47
don't mind if the module gets obsoleted in a day.
Tene What are you needing user-defined ops for? I don't get what you're doing.
masak I need it for a number of modules. nothing Web.pm-specific. 06:49
to have a really good Set module, we need user-defined ops.
Tene nodnod
but what are you trying to parse them for if you can't run them? 06:50
masak Tene: if I can preparse a program, turn the user-defined ops into multisubs, and then run the modified program, I'll have user-defined ops before everyone else :) 06:54
I call the idea 'fudo' -- Fake User-Defined Operators. 06:55
Tene okay, I can successfully match against Perl6::Grammar::TOP 07:02
but I can't get a Match object that I can investigate
moritz_ rakudo: say (1 ~~ /<Perl6::Grammar::TOP>).perl 07:13
p6eval rakudo 0d5515: OUTPUT«say requires an argument at line 1, near " (1 ~~ /<P"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:85)␤»
moritz_ rakudo: say (1 ~~ /<Perl6::Grammar::TOP>/).perl 07:14
p6eval rakudo 0d5515: OUTPUT«No result object␤current instr.: 'parrot;PCT;Grammar;ast' pc 86 (src/PCT/Grammar.pir:87)␤»
Tene rakudo: say (1 ~~ /<Perl6::Grammar::TOP>/)<perl><statement_block><statementlist><statement>[0]; 07:17
p6eval rakudo 0d5515: OUTPUT«Method 'postcircumfix:{ }' not found for invocant of class 'Failure'␤current instr.: 'postcircumfix:{ }' pc 5633 (src/classes/Associative.pir:133)␤»
moritz_ Tene: the top level capture would be Perl6::Grammar::TOP 07:18
Tene: unless you write Perl6::Grammar.parse(...)
Tene rakudo: 'say 1' ~~ /<Perl6::Grammar::TOP>/;say $/<Perl6::Grammar::TOP><statement_block><statementlist><statement>[0];
p6eval rakudo 0d5515: OUTPUT«say 1␤»
Tene now, if I can just get the positional and named...
eh, i guess that works... 07:23
Q:PIR
07:27 szabgab joined
masak yuck. :) 07:38
ihrd woow, this is wonderfull ( I readed your discussion), use rakudo grammar to colorize code will be much faster 07:52
masak ihrd: it is, yes. 07:53
should be able to parse the stuff output by --target=parse. 07:56
hm, or maybe run it through a simple filter to make it valid Perl 6. 07:57
it almost is already.
07:58 wayland__ joined
masak I played around with filtering line-by-line using regexes. conclusion: probably a better idea to build a grammar. 08:08
wayland76 I presume you know that azawawi did a HTML Perl6 colouriser? 10:33
12:27 ruoso joined
masak wayland76: aye. 12:36
zarah masak: You have new messages. Write '/msg zarah @messages' to read them.
masak wayland76: that might be good to look at, even though azawawi used Perl 5 for is colourizer. 14:26
19:03 sri_kraih joined 22:14 mj41_ joined 22:19 moritz_ joined 22:24 Matt-W joined 23:19 ihrd joined