[MERGE] TreeTransform avoids many renames when constructing trees

[Aaron Bentley]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

John Arbash Meinel wrote:

> What I found was that for long (100k entry) lists that:
> 
> foo = sorted([x for x in something])
> 
> is slightly (<10%) faster than:
> 
> foo = sorted(x for x in something)
> 
> My only guess is that it can pre-allocate somehow. The tradeoff, however, is
> that you have another copy of the data in memory. (in the temporary list).

It's also possible that it's because using a generator expression is
slower than using a list comprehension.  That seems perfectly reasonable
to me.  Iterators are not orders of magnitude faster than function calls.

> can do the same thing at effectively the same speed but without creating an
> intermediate list.
> 
> I also just found that doing:
> 
> x = [n for n in y]
> x.sort()
> 
> isn't strictly slower than
> x = sorted([n for n in y])

Well, that's an odd result.

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGZY340F+nu1YWqI0RAqqwAJ9lUEo7VncbRxbRk1sLGm4sRdY5HQCaA6Vd
lPWL/E7qyz0XbY1o0b8pyfM=
=63YW
-----END PGP SIGNATURE-----