Is it just Syntactic Sugar ?

[Johann Hibschman]

Michael Hudson writes:
> Johann Hibschman <johann at physics.berkeley.edu> writes:

>> But then it loses its most basic utility, namely working with
>> integers.  An integer is not a mutable type.
>> 
>> A simple rewrite is the only way to make sense of
>> 
>> i += 1  -> i = i + 1
>> 
>> if i is an integer, given the existing python variable semantics.

> But it's not the only way to make sense of

> a.b.c[item(d)] += e

> which you'd like to be roughly equivalent to

> t1 = a.b.c
> t2 = item(d)
> t1[t2] = t1[t2] + e

> or 

> a.b.c.__add_item__(item(d),e)

> depending on whether a.b.c is an instance that implement the
> appropriate magic or not.

Ah, I see.  That's true.  I would rather see a macro-like definition
as you outlined first, with some python equivalent of gensym.  Yes.
Hmm.

I still see this as a macro-like method.  And, yes, what you've
outlined does seem like the proper way to expand the macro, will full
gensym-ing and all.  But it still calls the simple "+" operator on the
object returned by t1[t2], not some form of t1[t2].incr(e), which is
what I'd understood Thomas Wouters to be asking for.

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu