bug#73431: Add `setf` support for `stream.el` in ELPA

[Okamsn via "Bug reports for GNU Emacs, the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>]

Stefan Monnier wrote:
>>> +(defun \(setf\ stream-first\) (store stream)
>>> +  "Set the first element of STREAM to value STORE."
>>> +  (if (stream-empty-p stream)
>>> +      (error "Cannot set first element of empty stream: %s" stream)
>>> +    (setcar (stream--force stream) store)))
>>
>> I am not sure what the preferred practice to define generalised setters
>> is.  In gv.el everything is defined using `gv-define-simple-setter' or
>> `gv-define-setter', which /feels/ more robust?  I believe that Stefan
>> (as the author or gv.el) might be able to explain if this is so or not.
> 
> Defining \(setf\ FOO\) looks fine to me 🙂
> I'm not sure we want to make streams mutable, OTOH.
> Is there a known use-case for it?
> 
> 
>          Stefan
> 

Hello,

Currently, using `(setf (seq-elt STREAM 0) VAL)` silently fails, because 
it treats the stream as a list, breaking the stream.

On the desire for mutability, there is the included macro `stream-pop`.

My use case is mainly consistency. I am currently cleaning up support 
for destructuring generic sequences with generalized variables in my 
package, which is how I noticed the silent failure for streams. I have 
found streams useful for iterating over sub-sequences of vectors, like 
what `cl-maplist` does with lists.

Thank you.