Python: *expression

Yao Yao on September 25, 2016

Argument Unpacking: *expression and **expression inside function calls

From The Python Language Reference - 6.3.4. Calls:

If the syntax *expression appears in the function call, expression must evaluate to an iterable. Elements from these iterables are treated as if they were additional positional arguments. For the call f(x1, x2, *y, x3, x4), if y evaluates to a sequence y1, ..., yM, this is equivalent to a call f(x1, x2, y1, ..., yM, x3, x4).

If the syntax **expression appears in the function call, expression must evaluate to a mapping, the contents of which are treated as additional keyword arguments. If a keyword is already present (as an explicit keyword argument, or from another unpacking), a TypeError exception is raised.

Pay attention to the parameter order: *expression syntax can appear after explicit keyword arguments, and be processed prior to the keyword arguments and any **expression arguments. E.g.

def func(a, b ,c):
    print(a, b, c)

func(c=3, 1, 2)                # SyntaxError: positional argument follows keyword argument
func(c=3, *(1,2))              # OK. 1 2 3
func(c=3, **dict(a=1,b=2))     # OK. 1 2 3
func(c=3, *(1,), **dict(b=2))  # OK. 1 2 3

Parameter Packing: *expression and **expression in function definitions

On the other hand, from The Python Language Reference - 8.6. Function definitions:

If the form *identifier is present, it is initialized to a tuple receiving any excess positional parameters, defaulting to the empty tuple.

If the form **identifier is present, it is initialized to a new dictionary receiving any excess keyword arguments, defaulting to a new empty dictionary.

  • So *args is definitely a tuple!

Let’s construct an example:

def func(*args, **kwargs):

func(1, a=2)

# output:
#   (1,)
#   {'a': 2}
#   {'kwargs': {'a': 2}, 'args': (1,)}

The syntax of one-elemented tuples is kind of weird. Just get used to it. The Python Language Reference - 6.14. Expression lists indicates:

The trailing comma is required only to create a single tuple (a.k.a. a singleton); it is optional in all other cases. A single expression without a trailing comma doesn’t create a tuple, but rather yields the value of that expression.

Unpacking inside tuple, list, set and dictionary displays

  • N.B. This PEP does NOT specify unpacking operators inside list, set or dictionary comprehensions.

From PEP 448 – Additional Unpacking Generalizations:

>>> *range(4), 4
(0, 1, 2, 3, 4)
>>> [*range(4), 4]
[0, 1, 2, 3, 4]
>>> {*range(4), 4}
{0, 1, 2, 3, 4}
>>> {'x': 1, **{'y': 2}}
{'x': 1, 'y': 2}

In dictionaries, latter values will always override former ones:

>>> {'x': 1, **{'x': 2}}
{'x': 2}
>>> {**{'x': 2}, 'x': 1}
{'x': 1}
  • N.B. We can also call
    • *: iterable unpacking operator and
    • **: dictionary unpacking operator

Extended Unpacking: *expression on LHS of assignments

From PEP 3132 – Extended Iterable Unpacking:

A tuple (or list) on the left side of a simple assignment may contain at most one expression prepended with a single asterisk (which is henceforth called a “starred” expression, while the other expressions in the list are called “mandatory”).

  • Mandatory expressions will be assigned the corresponding values of RHS according to their positions in the tuple (or list)
  • The starred expression will catch the remainder values of RHS

E.g. if seq is a slicable sequence, all the following assignments are equivalent if seq has at least 2 elements:

a, *b, c = seq
[a, *b, c] = seq
a, b, c = seq[0], list(seq[1:-1]), seq[-1]
  • seq[0] is guaranteed to be assigned to a
  • seq[-1] is guaranteed to be assigned to c
  • All the remainder values in seq will be assigned to b
    • b is always a list as far as I experimented
  • If len(seq) == 2, b will be empty

It is also an error to use the starred expression as a lone assignment target, as in

*a = range(5)  # Error

This, however, is valid syntax:

*a, = range(5)  # OK

This proposal also applies to tuples in implicit assignment context, such as in a for statement:

for a, *b in [(1, 2, 3), (4, 5, 6, 7)]:
# output:
#   [2, 3]
#   [5, 6, 7]

More examples in stack overflow: Unpacking, Extended unpacking, and nested extended unpacking.

blog comments powered by Disqus