编写Python2与Python3兼容代码：__future__和future如何选择？

Great question—setting up solid compatibility habits early is smart, especially since your code will grow more complex later. Let’s break down the specific scenarios where each tool shines, based on your needs.

When to Use __future__ #

__future__ is a built-in Python module that lets you enable Python 3 syntax and behavior in Python 2 ahead of time. It’s part of the core language, so no extra installation is needed.

• Syntax-level compatibility: Use it when you want Py2 to behave like Py3 for fundamental language features. For example:
  • from __future__ import print_function: Forces Py2 to use the Py3-style print() function instead of the print statement (critical for your current use case!).
  • from __future__ import division: Makes / return a float (true division) in Py2, matching Py3, instead of integer division.
  • from __future__ import unicode_literals: Makes all string literals ("hello") default to unicode in Py2, aligning with Py3’s str type.
• No extra dependencies: If you want to avoid adding third-party packages to your project initially, __future__ is perfect—it’s baked into Python 2.6+ and 3.3+ (which matches your supported version range).
• Simple, foundational changes: For basic code (like your current print-only setup), __future__ gives you the most essential compatibility wins without overcomplicating things.

When to Use the future Module #

The future module is a third-party library (install via pip install future) that provides a broader compatibility layer, focusing on standard library parity and advanced Py3 features in Py2.

• Standard library unification: Use it when you want to write code that uses Py3’s standard library structure in both versions. For example:
  After running from future.standard_library import install_aliases; install_aliases(), you can write import urllib.request instead of juggling Py2’s urllib2 and Py3’s urllib.request with conditional imports.
• Access Py3-only features in Py2: If your future code needs features like pathlib, enum, or concurrent.futures that don’t exist natively in Py2, future provides backported implementations.
• Reduce boilerplate: As your code grows, future eliminates the need for messy version checks (like if sys.version_info >= (3,0): ... else: ...) for many common tasks. It’s especially useful if you want to write code that feels "native" to Py3 while still supporting Py2.

Recommendations for Your Project #

Since your code is simple now but will get more complex later, here’s a step-by-step approach:

1. Start with __future__: Add these imports at the top of every file immediately:

   from __future__ import print_function, division, unicode_literals
   

   This fixes the most common Py2/Py3 mismatches and sets a solid foundation.
2. Introduce future when needed: Once you start using standard library modules that differ between versions (like urllib, subprocess, or io) or need Py3 features in Py2, install future and use its tools to unify your code.
3. Optional: Pair with six: For even more granular compatibility (like handling class inheritance or iterator changes), six works well alongside __future__ and future, but it’s not strictly necessary until your code gets very complex.

内容的提问来源于stack exchange，提问作者Lolo