default behavior

[Wolfram Hinderer]

After some off-list discussion with Ethan Furman (many thanks!), the
Python Doc bug is submitted: #9536 at bugs.python.org.

-John

This is probably nitpicking, but the patch calls __missing__ a special
method. However, unlike special methods, it is not invoked by "special
syntax" but by the dict's __getitem__ method. (len() invokes __len__
on any object - you can't do something similar with __missing__.)

__missing__ is also not listed as a special method on
http://docs.python.org/py3k/reference/datamodel.html#special-method-names

However, "normal" special method lookup seems to be used.

 

[John Posner]

This is probably nitpicking, but the patch calls __missing__ a special
method. However, unlike special methods, it is not invoked by "special
syntax" but by the dict's __getitem__ method. (len() invokes __len__
on any object - you can't do something similar with __missing__.)

__missing__ is also not listed as a special method on
http://docs.python.org/py3k/reference/datamodel.html#special-method-names

However, "normal" special method lookup seems to be used.

Fair enough. Please add your comment to #9536 at bugs.python.org.

Tx,
John

 

[Stefan Schwarzer]

On Fri, 30 Jul 2010 08:34:52 -0400, wheres pythonmonks wrote:
It does re-use the same underlying data.

from collections import defaultdict as dd
x = dd(list)
x[1].append(1)
x

y = dict(x)
x[1].append(42)
y

{1: [1, 42]}

One thing to keep in mind: dict(some_defaultdict) doesn't
store a reference to the defaultdict; instead it makes a
shallow copy, so key/value pairs added _after_ the "cast"
aren't included in the new dict:

y[2] = 17
y {1: [1, 42], 2: 17}
x

defaultdict(<type 'list'>, {1: [1, 42]})

Stefan

 

[David Niergarth]

This is a syntax I never noticed before. My built-in complier (eyes)
took one look and said: "that doesn't work." Has this always worked in
Python but I never noticed? I see other instance examples also work.
True

and properties
1.0

Curiously, this works

 

[David Niergarth]

[Oops, now complete...]
This is a syntax I never noticed before. My built-in complier (eyes)
took one look and said: "that doesn't work." Has this always worked in
Python but I never noticed? I see other instance examples also work.

  >>> '1' .zfill(2)
  '01'
  >>> 1.0 .is_integer()
  True

and properties

  >>> 1.0 .real
  1.0

Curiously, a float literal works without space
1.0

but not an int.
File "<stdin>", line 1
1.conjugate()
^
SyntaxError: invalid syntax

Anyway, I didn't realize int has a method you can call.

--David

 

[Peter Otten]

[Oops, now complete...]
This is a syntax I never noticed before. My built-in complier (eyes)
took one look and said: "that doesn't work."

(1).conjugate may hurt a little less. Anyway, the space is only needed for
the tokenizer that without it would produce a float immediately followed by
a name.

 

[Steven D'Aprano]

This is a syntax I never noticed before. My built-in complier (eyes)
took one look and said: "that doesn't work." Has this always worked in
Python but I never noticed?

Yes. Here is is working in Python 2.2:

[steve@sylar ~]$ python2.2
Python 2.2.3 (#1, Aug 12 2010, 01:08:27)
[GCC 4.1.2 20070925 (Red Hat 4.1.2-27)] on linux2
Type "help", "copyright", "credits" or "license" for more information.5


Syntactically, it also worked as far back as Python 1.5, although it is
rather pointless since int objects didn't gain any methods until 2.2:

[steve@sylar ~]$ python1.5
Python 1.5.2 (#1, Apr 1 2009, 22:55:54) [GCC 4.1.2 20070925 (Red Hat
4.1.2-27)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, AmsterdamTraceback (innermost last):
File "<stdin>", line 1, in ?
AttributeError: 'int' object has no attribute '__add__'

I see other instance examples also work.

'01'

You don't need the space between strings and the attribute access:
"1".zfill(2) is fine. You only need it for numbers, due to the ambiguity
between the decimal point and dotted attribute access.

 

[Piet van Oostrum]

You don't need the space between strings and the attribute access:
"1".zfill(2) is fine. You only need it for numbers, due to the ambiguity
between the decimal point and dotted attribute access.

Personally I prefer parentheses: (1).conjugate