🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Logs available at irclogs.raku.org/raku-dev/live.html | For MoarVM see #moarvm
Set by lizmat on 8 June 2022.
00:01 Voldenet left 00:21 Voldenet joined 00:35 buildable6 left 00:36 buildable6 joined 01:36 buildable6 left 01:39 buildable6 joined 01:51 buildable6__ joined, buildable6 left 02:39 buildable6__ left 02:43 buildable6 joined 03:43 buildable6 left 03:45 buildable6 joined 03:49 buildable6 left, buildable6 joined 04:45 buildable6 left 04:48 buildable6 joined 05:48 buildable6 left 05:50 buildable6 joined 06:50 buildable6 left 06:53 buildable6 joined
Geth rakudo/main: 53a79941f9 | (Christian Bartolomäus)++ (committed using GitHub Web editor) | src/core.c/Backtrace.rakumod
(Re-)align backtraces on JVM with MoarVM (#5421)

With 04a3127569 backtraces on the JVM backend started to include references to NQP's internals. This commit intends to apply the changes from 04a3127569 to the JVM backend, too.
07:53 buildable6 left 07:55 buildable6 joined 08:25 sena_kun joined 08:54 sjn left 08:55 buildable6 left 08:56 buildable6 joined 08:58 sjn joined
Geth rakudo/main: bd5a28c904 | (Elizabeth Mattijsen)++ | 18 files
RakuAST: add localization of quote-lang markers

Specifically: m ms q Q qq rx s S ss Ss .
Technically, ms ss Ss are really synonyms for m:ss s:ss S:ss respectively but that proved very hard to solve with the given grammar tokens, because all of these markers need to check whether there is an entity ... (10 more lines)
09:56 buildable6 left 10:00 buildable6 joined
Geth rakudo/main: 05a2c26ef6 | (Elizabeth Mattijsen)++ | 3 files
RakuAST: add missing entries to French localization
rakudo/main: d8d79a323c | (Elizabeth Mattijsen)++ | 5 files
RakuAST: "ss" is not a quoting adverb
11:00 buildable6 left 11:06 buildable6 joined 11:10 buildable6 left, buildable6 joined 11:38 sena_kun left
nine It's becoming quite obvious that post processing the QAST for BEGIN time execution is actually the only way to go. 11:41
There are cases where we generate the QAST before we know that we will execute it at BEGIN time. And there are cases where we execute at BEGIN time and also at runtime which will require different QAST. 11:42
This again leaves the question of where to get the static values from when I replace QAST::Vars in post procesing with QAST::WVals. 11:43
I think, I may be able to annotate the QAST::Vars with their originating RakuAST nodes. That would give me access to the resolved values.
lizmat why not annotate them with the constant value to begin with ? 11:47
ah, too soon I guess ok
nine Well....that is actually not a bad question at all. 11:49
This could cause some unwanted early processing. The value may e.g. be the meta object. Though I guess we rarely make it through QAST creation without also creating meta objects. 11:50
lizmat unwanted or unneeded ? 11:51
Geth rakudo/main: 36844606d8 | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: fix issue with < > in regexes

Turns out we should simply use $*LANG as the invocant to get the appropriate quote language inside a regex
nine Both I guess :) Unneeded since we should be able to defer getting the resolved value and unwanted because early anything usually triggers the need to do other things earlier. 11:57
Geth rakudo/main: 2b365ac9e6 | (Elizabeth Mattijsen)++ | 3 files
Rakuast: fix NL localization on qq//

It's dq now. Also use as a test for the functionality of localizing quote-lang markers.
12:06 buildable6 left
lizmat m: $_ = "bar"; say m/ < foo bar >/ # so this works 12:07
camelia 「bar」
lizmat m: $_ = "bar"; say m/ << foo bar >>/
camelia Potential difficulties:
Space is not significant here; please use quotes or :s (:sigspace) modifier (or, to suppress this warning, omit the space, or otherwise change the spacing)
at <tmp>:1
------> $_ = "bar"; say m/ << foo⏏…
lizmat so why doesn't that ?
12:09 buildable6 joined
ab5tract_ m: Q| $_ = “bar”; say m/<< foo bar >>/ |.AST.EVAL 12:09
camelia Potential difficulties:
Space is not significant here; please use quotes or :s (:sigspace)
modifier (or, to suppress this warning, omit the space, or otherwise
change the spacing).
------> $_ = “bar”; say m/<< foo⏏…
ab5tract_ intriguing 12:10
lizmat yeah, the fix should be easy in the Raku grammar, and I have half a mind to just implement it :-) 12:11
but after some fresh air&
13:09 buildable6 left 13:10 buildable6 joined 14:10 buildable6 left 14:12 buildable6 joined
Geth nqp/main: 15ad295980 | (Christian Bartolomäus)++ | 2 files
Tweak docs for continuations (s/shift/control/)

The remaining references to continuationshift were overlooked back in 2013 in 6cfee5accc (following up on 9aeeadd231).
  (Also replace ".perl" with ".raku" in the example code.)
rakudo: vrurg++ created pull request #5439:
Fix handling of dynamic containers on 6.e PseudoStash
15:12 buildable6 left
Geth rakudo/main: 0479e2fd8b | (Vadim Belman)++ | src/core.e/PseudoStash.rakumod
Fix handling of dynamic containers on 6.e PseudoStash
Before it was trying to provide support for Array and Hash containers.

container the code was considering as dynamic symbols any kind of symbol which has a returning true `dynamic` method. Aside of that, attempt to test for the method on roles was causing them to pun unintentionally.
rakudo/main: 663da63ae6 | (Vadim Belman)++ (committed using GitHub Web editor) | src/core.e/PseudoStash.rakumod
Merge pull request #5439 from vrurg/fix-pseudostash-container-handling

Fix handling of dynamic containers on 6.e PseudoStash
15:21 buildable6 joined 15:25 buildable6 left, buildable6 joined
lizmat m: $_ = "FOO"; say m/ :5i foo / # /me wonders what the 5 means :-) 16:04
camelia 「FOO」
lizmat m: $_ = "FOO"; say m/ :i(42) foo / 16:11
camelia 「FOO」
lizmat m: $_ = "FOO"; say m/ :i( 42) foo /
camelia Nil
lizmat weird 16:12
ab5tract_ Seriously strange 16:14
lizmat yeah, I'm not going to allow that nonsense in the Raku grammar 16:16
it's all from NQP's grammar QRegex::P6Regex::Grammar
ab5tract_ lizmat++
lizmat aahhh., I realize what the nonsense is about: allowing :dba(foo) 16:18
16:21 buildable6 left 16:22 buildable6 joined 17:15 sena_kun joined 17:22 buildable6 left 17:23 buildable6 joined 17:28 buildable6 left, buildable6 joined 17:32 buildable6 left, buildable6 joined
ab5tract_ I’ve been curious what that dba stuff is all about 17:45
lizmat it''s a form of comment 17:52
doing business as
why it is implemented as a internal modifier, beats me
I haven't seen any place it's being introspected internally
18:23 buildable6 left 18:25 buildable6 joined
Geth rakudo/main: 9a115ba4f4 | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: do quote adverb localization at the right place
rakudo/main: 7fcbfd657c | (Elizabeth Mattijsen)++ | 2 files
RakuAST: add localization of internal regex modifiers

  - hide token metachar:sym<mod> from NQP with a local implementation
  - create a local token metachar:sym<dba> for handling :dba()
  - use regex adverb lookup to do internal modifier localization
This removes support for the undocumented (and untested) :42i and :i(42) ... (7 more lines)
19:25 buildable6 left 19:28 buildable6 joined
ab5tract_ Seems that we had a regression in RakuAST: `t/12-rakuast/regex-charclass.rakutest` is failing for me with HEAD 20:08
wait, what 20:10
Ok, seems to be a PEBKAC 20:11
lizmat: I think there might be something broken about the way those tests are written. Why would $foo ~~ $_ return $foo? 20:21
m: $_ = “bar”; dd “foo” ~~ $_ 20:22
camelia Bool::False
20:28 buildable6 left
ab5tract_ Ahh, because it is a regex object… 20:28
damnit :(
20:30 buildable6 joined
ab5tract_ Okay, this looks broken to me: 20:35
m: /foo/ ~~ “foo” 20:36
camelia Regex object coerced to string (please use .gist or .raku to do that)
in block <unit> at <tmp> line 1
ab5tract_ Why doesn’t a string know how to ACCEPTS a regex without coercion? 20:38
nemokosch how would a string accept a regex? 21:08
ab5tract_ By calling $regex.ACCEPTS(self) 21:12
so yeah, not the right way to go 21:13
21:30 buildable6 left 21:32 buildable6 joined
Geth rakudo/main: 60f4b8b9fb | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: :dba is not a Raku thing

So don't bother looking for it
ab5tract_ :dba(“done doing business”) :) 22:03
22:29 sena_kun left 22:32 buildable6 left 22:34 buildable6 joined
Geth rakudo: ab5tract++ created pull request #5440:
RakuAST: Avoid re-declaration of `$_` in sub-sig
23:33 buildable6 left 23:34 buildable6 joined