00:16 lucasb left
jnthn lizmat: I don't do reddit, but quickly glancing it, I don't really get the issue; sockets don't do newline translation, and if you're writing tests, just make sure that you test against \r\n? Like Cro's tests do: github.com/croservices/cro-http/bl...rser.t#L15 00:16
Further, if you just put a `\r\n` straight into a string literal, it isn't translated 00:20
m: say "\r\n" ~~ /\r\n/
camelia
jnthn heh :)
m: dd "\r\n" ~~ /\r\n/
camelia Match.new(made => Any, hash => Map.new(()), orig => "\r\n", from => 0, pos => 1, list => ())
00:57 Hotbees joined
vrurg How do damn methods of a role preserve closure of a parameterized body block??? I'll get touched soon... 01:42
Xliff vrurg: o/ 02:39
vrurg Xliff: o/ 02:40
03:52 AlexDani` joined 03:56 AlexDaniel left 03:59 entonian joined 04:04 entonian left 05:04 squashable6 left, nativecallable6 left, statisfiable6 left, unicodable6 left, coverable6 left, notable6 left, greppable6 left, committable6 left, quotable6 left, reportable6 left, shareable6 left, benchable6 left, releasable6 left 05:05 bloatable6 left, greppable6 joined, coverable6 joined, reportable6 joined 05:06 squashable6 joined, bloatable6 joined, releasable6 joined, benchable6 joined, quotable6 joined, shareable6 joined 05:07 committable6 joined, notable6 joined, statisfiable6 joined, unicodable6 joined, nativecallable6 joined 07:34 sena_kun joined 08:15 [Tux] joined
sena_kun github.com/rakudo/rakudo/issues/2554 can be closed now. \o/ 08:23
tellable6 2019-11-12T00:04:42Z #raku <guifa> sena_kun :-D llunes, 12 de payares de 2019 a les 00:02:36; 2019年11月12日月曜日 0時02分36秒; понеделник, 12 ноември 2019 г., 0:04:02 ч.
lizmat m: dd "foo" ~~ / < [fo] > / # so, what is this supposed to match ? 09:00
camelia Nil
lizmat a token " [fo] " ?
09:00 robertle joined 09:08 Altai-man_ joined 09:10 sena_kun left
lizmat m: say "foo" ~~ / <?after f> o / 09:10
camelia 「o」
lizmat m: say "foo" ~~ / <after f> o /
camelia 「o」
after #<failed match>
lizmat so, what is the difference between "<?after" and "<after" ? 09:11
(yes, I'm working on writing a chapter on the differences in regular expressions between Perl and Raku)
moritz <?after doesn't capture 09:25
the ? turns anything into a zero-width assertion, and stops capturing. Since "after" already is zero-width, the effect isn't big here 09:26
lizmat but why the difference in output ?
moritz m: "foo" ~~ / <after f> o /; say $<after>.from; say $<after>.to 09:28
camelia 2
1
moritz lizmat: the difference in output is because <after> captures 09:30
and the $<after> match stringification is questionable
background: NQP match objects use the criteria $.from > $.to to mark failed matches
lizmat ah... ok 09:31
moritz but I think the bug is that $<after>.from > $<after>.to
probably related to look-behinds working on reversed strings or something
m: "foo" ~~ / <after f> o /; say $<after>.Str 09:32
camelia
lizmat so gistification is more to blame ? 09:33
moritz no, I'd argue that it's wrong that $<after>.to and $<after>.from are reversed 09:34
and the weird .gist is a result of that 09:36
and I think .Str should say "f" not ""
lizmat ok, need to go afk now 09:37
will make an issue for this, unless someone beats me to it :-)
10:00 epony joined 10:20 AlexDani` is now known as AlexDaniel, AlexDaniel left, AlexDaniel joined
|Tux| Rakudo version 2019.07.1-478-g75056dfa1 - MoarVM version 2019.07.1-316-g23daabadc
csv-ip5xs0.695 - 0.733
csv-ip5xs-206.373 - 6.478
csv-parser20.924 - 21.652
csv-test-xs-200.423 - 0.429
test5.865 - 6.062
test-t1.773 - 1.788
test-t --race0.816 - 0.896
test-t-2028.560 - 30.529
test-t-20 --race9.125 - 9.285
10:35
10:56 Kaiepi left, Kaiepi joined 11:08 sena_kun joined 11:11 Altai-man_ left 12:52 cognomin_ left, tobs left, masak left, nine left, SyrupThinker left 12:53 cognomin_ joined, tobs joined, tyil joined, masak joined, SyrupThinker joined, nine joined 13:07 lucasb joined 13:08 Altai-man_ joined 13:11 sena_kun left
lizmat m: @*ARGS = "foo"; sub MAIN(Int:D $dial) { say “Frobnicator dial is at $dial!” } 13:55
camelia Usage:
<tmp> <dial>
lizmat in my mind, that should say that an Int is required, should it not ?
13:55 Kaiepi left 13:57 Kaiepi joined
lizmat looks like it does for named parameters: 14:01
m: sub MAIN(Int:D :$dial) { say “Frobnicator dial is at $dial!” }
camelia Usage:
<tmp> [--dial=<Int>]
14:07 robertle left 14:08 Kaiepi left 14:09 robertle joined, Kaiepi joined 15:08 sena_kun joined 15:10 Altai-man_ left 16:03 robertle left 16:20 Kaiepi left 16:40 ExtraCrispy joined
Geth rakudo/add-type-to-positional-in-usage: 360363df67 | (Elizabeth Mattijsen)++ | src/core.c/Main.pm6
Add any type information to positional in usage

Given a sub MAIN(Int $dial) in a script "frobnicate.p6", and calling it without a single integer value, would give:
   Usage:
   frobnicator.p6 <dial>
... (7 more lines)
16:45
rakudo: lizmat++ created pull request #3300:
Add any type information to positional in usage
16:59 robertle joined 17:08 Altai-man_ joined 17:10 sena_kun left 18:00 MasterDuke joined 18:21 Demos[m] left, rba[m] left, tyil[m] left, AlexDaniel` left, Kaiepi joined, timotimo[m] left 19:09 sena_kun joined 19:10 Altai-man_ left 19:29 ExtraCrispy left 19:45 MasterDuke left 20:21 MasterDuke joined 21:08 Altai-man_ joined 21:10 sena_kun left 21:27 lucasb left 22:38 Altai-man_ left 23:02 ZzZombo left 23:03 ZzZombo joined 23:33 cognominal joined 23:36 cognomin_ left