Doctest无法识别__future __。division

问题描述

我有以下doctest编写为 x.doctest ：

I have the following doctest written x.doctest:

This is something:

    >>> x = 3 + 4

foo bar something else:

    >>> from __future__ import division
    >>> y = 15
    >>> z = int('24')
    >>> m = z / y
    >>> print (m)
    1.6

但是当我运行 python时- m在python 2.7.11上的doctest x.doctest 中，doctest无法识别__future__进口部门的 ：

But when I ran python -m doctest x.doctest on python 2.7.11, the doctest didn't recognize from __future__ import division:

**********************************************************************
File "x.doctest", line 11, in x.doctest
Failed example:
    print (m)
Expected:
    1.6
Got:
    1
**********************************************************************
1 items had failures:
   1 of   6 in x.doctest
***Test Failed*** 1 failures.

即使我将未来的导入语句移到第一行：

Even when I shifted the future import statement to the first line:

This is something:

    >>> from __future__ import division
    >>> x = 3 + 4

foo bar something else:

    >>> y = 15
    >>> z = int('24')
    >>> m = z / y
    >>> print (m)
    1.6

doctest仍然失败：

The doctest still fails:

**********************************************************************
File "x.doctest", line 11, in x.doctest
Failed example:
    print (m)
Expected:
    1.6
Got:
    1
**********************************************************************
1 items had failures:
   1 of   6 in x.doctest
***Test Failed*** 1 failures.

为什么会这样，我该如何解决？

是否有doctest标志/选项，要求确保__future__进口部门的 被识别？

Is there a flag / option for doctest that asks ensures that from __future__ import division is recognized?

注意：我可以强制检查 print（int（m））或 y = 15。会通过doctest，但这不是很理想。

Note: I could just force the check on print (int(m)) or y = 15. and the doctest will pass but that is not that desirable.

推荐答案

Doctests通过Python编译器单独运行每行 。这意味着在doctest本身中用从__future__ import .. 语句声明的所有编译器标志都是无效的。

Doctests run each line in isolation through the Python compiler. This means that any compiler flags specified with a from __future__ import .. statement in the doctest itself is useless in a doctest.

但是，您可以从真实的到您的doctest全局变量。如果您不使用__future__ import< name> 格式的，但使用 import __future __ ，而是导入该实际模块，并可以将其定义的对象的引用添加到doctest globs 或 extraglobs 字典：

However, you can add names from the real __future__ module to your doctest globals. If you don't use the from __future__ import <name> format but use import __future__ instead, you import that actual module, and can add references to the objects it defines to the doctest globs or extraglobs dictionaries:

 if __name__ == "__main__":
     import doctest
     import __future__
     doctest.testmod(extraglobs={'division': __future__.division})

The <$ c然后，$ c> DocTestRunner 将为您设置正确的编译器标记，这些标记来自于其中。

The DocTestRunner will then set the right compiler flags for you when compiling individual lines from these.

演示：

>>> import doctest
>>> import __future__
>>> import sys
>>> def foo():
...     """
...     >>> 1 / 2
...     0.5
...     """
...
>>> doctest.testmod(sys.modules['__main__'])
**********************************************************************
File "__main__", line 3, in __main__.foo
Failed example:
    1 / 2
Expected:
    0.5
Got:
    0
**********************************************************************
1 items had failures:
   1 of   1 in __main__.foo
***Test Failed*** 1 failures.
TestResults(failed=1, attempted=1)
>>> doctest.testmod(sys.modules['__main__'], extraglobs={'division': __future__.division})
TestResults(failed=0, attempted=1)

这篇关于Doctest无法识别__future __。division的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！