sys_pre_expand compiler warning (Re: Algorithmic lists)

[Robert Virding]

Ulf Wiger <etxuwig@REDACTED> writes:
>On Mon, 16 Oct 2000, Ulf Wiger wrote:
>
>>
>>Here's my first (and perhaps last) hack of sys_pre_expand.erl.
>>It handles at least my very simple test function with a list
>>comprehension that is fed a lazy list:
>
>BTW, when I compiled sys_pre_expand.erl, I got the following compiler
>warning:
>
>> erlc -W sys_pre_expand.erl
>/home/etxuwig/work/erlang/lazy/sys_pre_expand.erl:703: Warning:
>variable 'Es' shadowed in 'fun'
>
>The function in question looks like this:
>
>bin_expand_strings(Es0) ->
>    foldr(fun ({bin_element,Line,{string,_,S},default,default}, Es) ->
>		  foldr(fun (C, Es) ->
>
>[{bin_element,Line,{integer,Line,C},default,default}|Es]
>			end, Es, S);
>	      (E, Es) -> [E|Es]
>	  end, [], Es0).
>
>
>Now, it doesn't really seem like a bug, but wouldn't it be awfully
>easy to avoid the warning and resulting confusion? ;)

The reason that there is a warning here (and in funs when head 
variables shadow) is that this is a departure from standard Erlang 
practice of NOT shadowing.  For example in

    X = ...,
    case ... of
        {a,X} ->

the X in the pattern inside the case does not shadow the outer X.  
(Which is contrary to most other functional languages)  It was felt 
that it would be better to highlight this to avoid confusion.  
Otherwise people might expect that a head variable in a fun should 
match if it already bound.

I agree it is sometimes irretating when I KNOW what I am doing (which 
of course I always do :-), but this is in keeping with the philosophy o 
ftrying to keep things simple, consistent and easy to learn so we 
highlight inconsistencies.

	Robert

-- 
Robert Virding                          Tel: +46 (0)8 545 55 017
Alteon Web Systems                      Email: rv@REDACTED
S:t Eriksgatan 44                       WWW: http://www.bluetail.com/~rv
SE-112 34 Stockholm, SWEDEN
"Folk säger att jag inte bryr mig om någonting, men det skiter jag i".