Re: anonymous functions? anonymous variables

So, if I understand the suggested improvement,

function($k, $v){$k}

will be written as:

  function($k, _){$k}

Hmm... this actually saves us writing just one character and doesn't seem
like a huge improvement, or is there something else I am missing?

We could even have (in a future XPath version):

   function($k, ){$k}

or generally something like:

  function($p1, , $p3, , , $p6 ) { Expr($p1, $p3, $p6) }

and it seems even better to have named parameters and specify them by name
-- regardless of order.

If we don't have the ability to specify a name for each argument in the
function declaration, then we would need to identify a specific argument by
its position number, which, I agree, is not as convenient as referring to
these by name.

To summarize, it seems that providing the ability for named arguments is a
good solution to the problem. If I remember well, Dr. Kay had something
like this in his XML Prague 2020 paper.


Dimitre

On Sun, Aug 16, 2020 at 1:57 PM Ihe Onwuka <[email protected]> wrote:

>
>
> On Sun, Aug 16, 2020 at 1:43 PM Dimitre Novatchev <[email protected]>
> wrote:
>
>> > I have to write
>> >
>> >    function ($k, $v) {$v}
>> >
>> >    instead of something like
>> >
>> >    function (_,$v) {$v}
>>
>> What is the rationale to have the signature of the function contain an
>> unreferenced argument? Why not use just:
>>
>>   function ($v)  { (: Expr :)}
>>
>>
> It may be a constraint imposed by the context in which the function is
> being used. See for example
> https://www.w3.org/TR/xslt-30/#func-map-for-each where the function
> supplied must have two arguments whereas the use case only requires the use
> of one.
>
>
>

_______________________________________________
saxon-help mailing list archived at http://saxon.markmail.org/
[email protected]
https://lists.sourceforge.net/lists/listinfo/saxon-help