As many of You know, python is such a strange language, that releases are not backward compatible. Ok, it's really not that bad, the problem only occurs when migrating from python 2.X to 3.X. Features introduced in python 3 greatly increase the interpreter performance (such as decimal number processing, function annotations) and provide other useful mechanisms (ex. exception chaining). Some of this things were technically not possible to achieve in python 2.X, so thats where the conflicts come from.
So if python 3 is that awesome, why do people still use python 2.7 (or older versions)? Most existing python frameworks and libraries do not support python 3, they are being ported but it will take some time for the whole python community to move on to the currently proposed version.
So if you plan migrating to python 3 in the future, you best use the __future__ module. This module provides some basic forward compatibility for older versions of python. Two major differences that may crash your code are the print statement and literals encoding. You may observe __future__ below, first the print function:
>>> print "test"
test
>>> from __future__ import print_function
>>> print "test"
File "<stdin>", line 1
print "test"
^
SyntaxError: invalid syntax
>>> print("test")
test
>>>
After importing print_function you are obligated to use print as a function. And now for unicode literals:
>>> isinstance("test", str)
True
>>> isinstance("test", unicode)
False
>>> from __future__ import unicode_literals
>>> isinstance("test", str)
False
>>> isinstance("test", unicode)
True
>>> "test"
u'test'
>>>
All string are now by default encoded as unicode... personally I'm happy with it, since I really hated the "Ascii codec can't decode..." errors.
No comments:
Post a Comment