[perl #118231] Importing the same thing twice shouldn't bomb

[via RT]

# New Ticket Created by  Elizabeth Mattijsen 
# Please include the string:  [perl #118231]
# in the subject line of all future correspondence about this issue. 
# <URL: https://rt.perl.org:443/rt3/Ticket/Display.html?id=118231 >


[23:47:51] <lizmat>      given a file A.pm with: module A:version<v1> { sub a 
is export { say "a" } }; module B:version<v1> { sub b is export { say "b" } }
[23:48:26]      stevan_ (~stev...@c-75-69-127-112.hsd1.ma.comcast.net) left 
IRC. (Remote host closed the connection)
[23:48:52] <lizmat>      use A; a; b; import B  # currently fails because it 
tries to import "b", whereas the "use A" already did that also
[23:49:32] <lizmat>      even though they are the same sub
[23:50:04]      sqirrel (~sqir...@brln-4db94d7a.pool.mediaways.net) left IRC. 
(Quit: Verlassend)
[23:50:12] <lizmat>      is the fact that "use A" exports the subs from module 
B a rakudobug, or is that a case of "don't do that" ?
[23:50:46] <jnthn>       Well, you're importing the stuff from EXPORT::DEFAULT
[23:50:52] <jnthn>       And "is export" installs the things in there
[23:51:01] <lizmat>      Feels to me that we should maybe disallow more than 
one module / class etc. at the top level of a UNIT to prevent these type of 
situations?
[23:51:11] <jnthn>       Nah.
[23:51:31] <jnthn>       It's common to do that. Especially if one of them is a 
lexical class that you're using as an implementation detail
[23:51:46] <lizmat>      I'm not against that
[23:51:57] <lizmat>      that's lexical inside, and hidden from the outside
[23:52:31] <lizmat>      it's just that there is a difference between "use A" 
and "import A" in the case of 2 top level modules with exportable subs
[23:52:32] <jnthn>       I thought we didn't complain about importing the same 
things twice, though...
[23:52:35] <jnthn>       r: use Test; use Test;
[23:52:36] <+camelia>    rakudo 25a0b3: 
OUTPUT«===SORRY!===␤Cannot import the following symbols from 
Test, because they already exist in this lexical scope:  &skip_rest, &diag, 
&done_testing, &done␤at /tmp/zuyxQPOgEB:1␤------> use Test; use 
Test⏏;␤»
[23:52:46] <jnthn>       Ah, hm.
[23:52:59] <lizmat>      so maybe that's the rakudobug ?
[23:53:10] <jnthn>       I'm a bit surprised it complains there.
[23:53:19] <jnthn>       Especially as I thought I'd fixed that in the past...
[23:53:43] <lizmat>      ok, then I'll submit a rakudobug for this and we can 
look at this later
[23:53:52] <lizmat>      is that a plan?
[23:54:02] <jnthn>       Sure