嵌套文档字符串的Doctest

Question

假设我有以下代码：

def foo(s):
    """A dummy function foo. For example:

>>> a = '''This is a test string line 1
This is a test string line 2
This is a test string line 3'''
>>> foo(a)
This is a test string line 1
This is a test string line 2
This is a test string line 3
>>>
    """
    print s

if __name__ == '__main__':
    import doctest
    doctest.testmod()

我们把它保存为 foo.py。当我运行：

C:\Python27>python.exe foo.py
**********************************************************************
File "foo.py", line 5, in __main__.foo
Failed example:
    a = '''This is a test string line 1
Exception raised:
    Traceback (most recent call last):
      File "C:\Python27\lib\doctest.py", line 1254, in __run
        compileflags, 1) in test.globs
      File "<doctest __main__.foo[0]>", line 1
        a = '''This is a test string line 1
                                          ^
    SyntaxError: EOF while scanning triple-quoted string literal
**********************************************************************
File "foo.py", line 8, in __main__.foo
Failed example:
    foo(a)
Exception raised:
    Traceback (most recent call last):
      File "C:\Python27\lib\doctest.py", line 1254, in __run
        compileflags, 1) in test.globs
      File "<doctest __main__.foo[1]>", line 1, in <module>
        foo(a)
    NameError: name 'a' is not defined
**********************************************************************
1 items had failures:
   2 of   2 in __main__.foo
***Test Failed*** 2 failures.

我尝试过给文档字符串缩进（比如说 >>> a = '''...'''，我检查了所有的缩进 - 每次缩进都是4个空格），还把单引号换成了双引号（>>> a = """...."""），但是错误信息不一样，文档测试就是无法成功。目前唯一有效的方法是把所有行合并成一个超长的字符串，并用 '\r\n' 来分隔。

我是不是漏掉了什么？

Answer 1

我觉得你需要在那儿加一些点。

>>> a = """This is a test string line 1
... This is a test string line 2
... This is a test string line 3"""