list in a tuple

[Carl Banks]

On Dec 27, 12:38 pm, montyphy... at gmail.com wrote:
> After some tought I must agree that this is a wart more than
> a bug and that it will probably be best not to mess with it.
> However, what do you guys think about the print wart in Py3k
> described athttp://filoxus.blogspot.com/2007/12/python-3000-how-mutable-is-immuta...
> (im not trying to advertise my blog, I just don't feel like
> typing the whole problem all over again)?


1. Tuples are immutable.  None of the tuples in your example were
modified.

The behavior you want (which is not immutability of tuples, which
Python already has, but *recursive* immutability of all objects
contained within the tuple) is not an unreasonable thing to ask for,
but the design of Python and common usage of tuples makes it all but
impossible at this point.

There is no general way to determine whether an object is mutable or
not.  (Python would have to add this capability, an extremely
substantial change, to grant your wish.  It won't happen.)

Tuples are used internally to represent the arguments of a function,
which are often mutable.

Tuples are sometimes used to return multiple values from a function,
which could include mutable values.

Tuples are used to specify multiple arguments to a format string, some
of which could be mutable, though I guess this is going away in Python
3.


2. The issue with print in your example is a bug, not a wart.  It'll
be fixed.

(This is just a guess, but I think it might have something to do with
the flux of the new bytes type.  The behavior manifested itself when
trying to print a self-referencing structure probably only because
that section of code was lagging behind.)


3. You're still top posting, which goes against this group's
conventions and annoys quite a few people.  When you reply to a
message, please move your cursor to below the quoted message before
you begin typing.  Thank you


Carl Banks