I would like to do something like this
def f():
return { \'a\' : 1, \'b\' : 2, \'c\' : 3 }
{ a, b } = f() # or { \'a\', \'b\' } = f() ?
<
mydict = {'aa':2, 'bb':'john', 'cc':34, 'dd':'bye'}
a, b, c, d = [mydict[k] for k in ['aa', 'bb', 'cc', 'dd']]
It wouldn't make any sense for unpacking to depend on the variable names. The closest you can get is:
a, b = [f()[k] for k in ('a', 'b')]
This, of course, evaluates f()
twice.
You could write a function:
def unpack(d, *keys)
return tuple(d[k] for k in keys)
Then do:
a, b = unpack(f(), 'a', 'b')
This is really all overkill though. Something simple would be better:
result = f()
a, b = result['a'], result['b']
Consider making f
a namedtuple
Then you can just use f.a
, f.b
directly
Hmm. Kind of odd since a dictionary is not ordered, so the value unpacking depends on the variable names. But, it's possible, if ugly:
>>> locals().update(f())
>>> a
1
Don't try this at home! It's a maintainability nightmare. But kinda cool too :-)
You could use (or abuse) a function attribute to do this:
>>> def f():
... f.a=1
... f.b=2
... return { 'a' : f.a, 'b' : f.b, 'c' : 3 }
...
>>> f()
{'a': 1, 'c': 3, 'b': 2}
>>> a,b=f.a,f.b
>>> a,b
(1, 2)
Be aware that the attributes only have value after f() is called or manually assigned.
I (rarely) use function attributes as a fill-in for C's static
class variables like so:
>>> def f(i):
... f.static+=i
... return f.static
>>> f.static=0
>>> f(1)
1
>>> f(3)
4
There are better ways to do this, but just to be complete...