probably worth a bug report ?

[Marc Chantreux]

hello rakoons,

I got this error message

Too few positionals passed; expected 1 argument but got 0
  in sub xxx at - line 1
  in block  at - line 2

Welcome to 𝐑𝐚𝐤𝐮𝐝𝐨™ v2021.09.
Implementing the 𝐑𝐚𝐤𝐮™ programming language v6.d.
Built on MoarVM version 2021.09.

by running this:

<<\-. raku; raku -v
sub xxx ( Pair $test ) { $test.raku.say }
xxx key => 'value';

and i think it's worth a bug: i don't know if

xxx key => 'value'

should parse but at least it deserve a less
confusing message, don't you think ?

regards
marc

Re: probably worth a bug report ?

[Elizabeth Mattijsen]

Maybe first explain why the error:

When you specify a Pair as such as an argument, it is interpreted as a *named* 
argument.


$ raku -e 'sub a(|c) { dd c }; a b => 42'
\(:b(42))

So thus you can see why the error is thrown: you passed 0 positional arguments 
to the subroutine, and it never gets to report the fact that you passed one 
unexpected named argument.

There's several ways to fix the problem:

xxx Pair.new("key","value");  # be explicit about sending a Pair

xxx (key => "value"); # add parentheses to break the syntactic 
sugar for named arguments

The latter option also shows that whitespace *can* be significant in Raku:

xxx(key => "value");   # same error

In Raku, an identifier directly followed by a parens open, means calling that 
object of that identifier, with the arguments specified.   And then you know 
that "xxx foo" is really syntactic sugar for "xxx(foo)".


Again, it is all about *reading* the error message.  Is there room for 
improvement?  There always is!

> On 2 Jan 2022, at 08:34, Marc Chantreux  wrote:
> 
> hello rakoons,
> 
> I got this error message
> 
>   Too few positionals passed; expected 1 argument but got 0
> in sub xxx at - line 1
> in block  at - line 2
>   
>   Welcome to 𝐑𝐚𝐤𝐮𝐝𝐨™ v2021.09.
>   Implementing the 𝐑𝐚𝐤𝐮™ programming language v6.d.
>   Built on MoarVM version 2021.09.
> 
> by running this:
> 
>   <<\-. raku; raku -v
>   sub xxx ( Pair $test ) { $test.raku.say }
>   xxx key => 'value';
> 
> and i think it's worth a bug: i don't know if
> 
>   xxx key => 'value'
> 
> should parse but at least it deserve a less
> confusing message, don't you think ?
> 
> regards
> marc

Re: probably worth a bug report ?

[Marc Chantreux]

Le Sun, Jan 02, 2022 at 12:32:46PM +0100, Elizabeth Mattijsen a écrit :
> Maybe first explain why the error

thanks for the explaination. especially

> $ raku -e 'sub a(|c) { dd c }; a b => 42'
> \(:b(42))

now my sub works the way I wanted:

sub got (|c) {
for c.hash.kv -> $rule ,$input {
ok (my $r = Pugish.subparse: $input , :$rule )
, "match a $rule in: $input";
$r and return $r;
}
}
got pair => 'obey the law';

> Again, it is all about *reading* the error message.
> Is there room for improvement?  There always is!

I wrote this message mostly because the error message didn't help at all
here but now i discovered and inspected the Capture, i realize the
message was true but just left me clueless.

Capture is a good way to understand what happened.

> > Too few positionals passed; expected 1 argument but got 0

because c.list got is all ...

many thanks for helping.
marc

Re: probably worth a bug report ?

[Ralph Mellor]

cf https://www.nntp.perl.org/group/perl.perl6.users/2021/04/msg9883.html

--
love, raiph

On Sun, Jan 2, 2022 at 7:35 AM Marc Chantreux  wrote:
>
> hello rakoons,
>
> I got this error message
>
> Too few positionals passed; expected 1 argument but got 0
>   in sub xxx at - line 1
>   in block  at - line 2
>
> Welcome to 𝐑𝐚𝐤𝐮𝐝𝐨™ v2021.09.
> Implementing the 𝐑𝐚𝐤𝐮™ programming language v6.d.
> Built on MoarVM version 2021.09.
>
> by running this:
>
> <<\-. raku; raku -v
> sub xxx ( Pair $test ) { $test.raku.say }
> xxx key => 'value';
>
> and i think it's worth a bug: i don't know if
>
> xxx key => 'value'
>
> should parse but at least it deserve a less
> confusing message, don't you think ?
>
> regards
> marc