🦋 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.
Geth rakudo/main: f151aba3c4 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: make sure Regex.source is populated

Although I'm not happy with assigning the .source at CHECK time, there should be a better way.
07:00
Geth whateverable: 9eff36756a | (Elizabeth Mattijsen)++ | 2 files
Fix issue with IRC::Client dependency

And bump version
08:38
Geth rakudo/main: 7114b4d739 | (Elizabeth Mattijsen)++ | 3 files
RakuAST: add set-from-also/from-also flag to Is/Does trait

So that we can mark the origin of the trait, needed for deparsing and .raku handling
12:06
Geth rakudo/main: 3072e91ea7 | (Elizabeth Mattijsen)++ | 3 files
Revert "RakuAST: add set-from-also/from-also flag to Is/Does trait"

This reverts commit 7114b4d739c1a51b5301859cf77ddc0af8716474.
Decided against creating a class for "also", as it opens up all sorts of issues with timing, deparsing and .raku
13:14
Geth rakudo/main: f325c40437 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: add support for "also is Foo" and "also does Foo"

Make this a purely grammar / action event, without any explicit deparsing or .raku. While 7114b4d739 introduced a flag for Traits to indicate they were from an "also", and while that would help in deparsing, it would not help with .raku.
... (16 more lines)
13:37
lizmat +3 tests!
vrurg_ lizmat++! 13:45
lizmat funny fact: "trusts" is implemented as a trait_mod, yet there is no grammar support for it as a trait 13:47
m: class B { ... }; class A trusts B { }
camelia ===SORRY!=== Error while compiling <tmp>
Unable to parse class definition
at <tmp>:1
------> class B { ... }; class A⏏ trusts B { }
expecting any of:
generic role
vrurg_ lizmat: technically, Perl6/Actions.nqp:3337 13:48
lizmat yes, but that sub could have been named anything else
vrurg_ I'd say the edges are fuzzy in this area. 13:50
Geth rakudo/main: e1f801fce7 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: add dummy .new to classes used as markers

Specifically ::Regex::Backtrack::xxx and ::Parameter::Slurpy::xxx
14:27
rakudo/main: 390b8227eb | (Elizabeth Mattijsen)++ | src/core.c/RakuAST/Raku.pm6
RakuAST: don't .rakufy :!excludes-min|max
14:33
ugexe why would an installed version of zef have problems deleting the home and site module locations? presumably its because some precomp or source file for zef (which would be located in one of those location) is open. but why would such a handle remain open after its loaded the module? 15:22
another instance such a problem occurs would be `zef install zef --force` with the same version of zef (which would generate the same sha1 file names and overwrite those source files)
Geth rakudo/main: 22e78454e1 | (Elizabeth Mattijsen)++ | src/core.c/RakuAST/Deparse.pm6
RakuAST: fix deparsing of ::Regex::Quantifier::(|Block)Range
15:43
Geth rakudo/main: 20bfeaf22e | (Elizabeth Mattijsen)++ | 2 files
RakuAST: also can also operate on subs/methods

So introduce a $*ALSO-TARGET that is set for packages and subs/methods and use that for handling "also"
16:27
Geth rakudo/main: 09b452bf8d | (Elizabeth Mattijsen)++ | 3 files
RakuAST: Streamline traits just taking a type

In all of these cases, the name of the trait was the only thing different. So create a RakuAST::Trait::Type class as a base class for the other RakuAST::Trait::xxx classes that only take a type as a positional. Adapt deparsing and rakufication accordingly.
18:40
rakudo/main: 1cf0b06d52 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: add support for ::Doc::Markup highlighting

And streamline use of it and some other tweaks
20:46
Geth rakudo/main: 4d5f6ff9e4 | (Nick Logan)++ | src/core.c/CompUnit/Repository/FileSystem.pm6
Deprecate including an ambiguous .pm module file

We need to know when raku source files have changed for the purposes of precompilation. However, when a user includes a library path
  (i.e. -Ilib, -I/home/foo) that does not contain a META6.json file
we might end up erroneously considering perl module files (which also use .pm) as raku files. While this is mostly harmless, it does waste some cpu and io. This adds a deprecation message when this ambiguous situation occurs so that in the future we might be able to remove support for the .pm extension altogether.
22:49
rakudo/main: 657dada537 | (Nick Logan)++ (committed using GitHub Web editor) | src/core.c/CompUnit/Repository/FileSystem.pm6
Merge pull request #5278 from rakudo/ugexe/deprecate-ambiguous-pm

Deprecate including an ambiguous .pm module file
[Coke] opens tickets for all the blin deps that trigger that warning... 23:40
ugexe I'm not sure any will 23:55
i guess modules included in t/ maybe
[Coke] hit like 5 or 6
ugexe but .pm files listed in META6.json are fine
i suppose anything that has a test with `use lib "lib"` would also do it 23:56
(that has .pm files in it)
[Coke] if, svg:plot, uri, shell::command, http useragent
file::directory:tree 23:57
ugexe http::useragent has .rakumod extension
oh yep, a test module
github.com/sergot/http-useragent/t...ster/t/lib
File::Directory::Tree has a `use lib "lib"` 23:58
in a test 23:59