28. Python脚本学习笔记二十八代码检测和分析

28. Python脚本学习笔记二十八代码检测和分析

本篇名言：“人生就象睡觉时盖了一条太短的毛毯,你往上拉,脚趾头抗议,往下拽,肩膀冻得哆嗦.只有乐观的人才会蜷起膝,美美地睡一晚.愿你天天快乐,夜夜好梦!”

除了直接测试外，其实还有其他工具来探索程序，用于代码检测和分析等。

代码检测是寻找代码中普通错误或者问题的方法，有点像编译器处理静态语言。

分析则是查明程序到底跑多快的方法。

单元测试让程序工作，源代码检查可以让程序更好，分析会让程序更快。

 

1.  PyChecker和PyLint

两个工具都是检查源代码。

1.1      Pylint下载

Pylint 处理pylint包之外，还需要astroid和logilab-common

http://www.logilab.org/projects/common

https://bitbucket.org/logilab/astroid

https://pypi.python.org/pypi/pylint/#downloads

                  蛤蟆已打包一并传到网上，链接如下：

http://download.csdn.net/detail/notbaron/8928061

解压后直接通过命令：python setup.py install

安装。

1.2      Pychecker下载安装

下载地址如下：http://pychecker.sourceforge.net/

执行如下检测：

E:\>pychecker  my_math

E:\python_zhizuo\Pythonlearn>D:\python27\python.exeD:\python27\Lib\site-package

s\pychecker\checker.py my_math

Processing module my_math (my_math)...

 

Warnings...

 

None

E:\>pylint my_math

No config file found, using defaultconfiguration

************* Module my_math

C: 4, 0: Exactly one space required around assignment

__revision__='0.1'

           ^ (bad-whitespace)

C: 5, 0: Exactly one space required after comma

def product(factor1,factor2):

                   ^ (bad-whitespace)

C: 7, 0: Final newline missing (missing-final-newline)

 

 

Report

======

4 statements analysed.

 

Statistics by type

------------------

 

+---------+-------+-----------+-----------+------------+---------+

|type    |number |old number |difference |%documented |%badname |

+=========+=======+===========+===========+============+=========+

|module  |1      |1          |=         |100.00      |0.00     |

+---------+-------+-----------+-----------+------------+---------+

|class   |0      |0          |=          |0           |0        |

+---------+-------+-----------+-----------+------------+---------+

|method  |0      |0          |=          |0           |0        |

+---------+-------+-----------+-----------+------------+---------+

|function |1      |1          |=          |100.00      |0.00    |

+---------+-------+-----------+-----------+------------+---------+

 

 

 

Raw metrics

-----------

 

+----------+-------+------+---------+-----------+

|type     |number |%     |previous|difference |

+==========+=======+======+=========+===========+

|code     |4      |44.44 |NC       |NC         |

+----------+-------+------+---------+-----------+

|docstring |4      |44.44 |NC       |NC         |

+----------+-------+------+---------+-----------+

|comment  |0      |0.00  |NC      |NC         |

+----------+-------+------+---------+-----------+

|empty    |1      |11.11 |NC       |NC        |

+----------+-------+------+---------+-----------+

 

 

 

Duplication

-----------

 

+-------------------------+------+---------+-----------+

|                         |now   |previous |difference |

+=========================+======+=========+===========+

|nb duplicated lines      |0    |0        |=          |

+-------------------------+------+---------+-----------+

|percent duplicated lines |0.000|0.000    |=          |

+-------------------------+------+---------+-----------+

 

 

 

Messages by category

--------------------

 

+-----------+-------+---------+-----------+

|type      |number |previous |difference |

+===========+=======+=========+===========+

|convention |3      |3       |=          |

+-----------+-------+---------+-----------+

|refactor  |0      |0        |=          |

+-----------+-------+---------+-----------+

|warning   |0      |0        |=          |

+-----------+-------+---------+-----------+

|error     |0      |0        |=          |

+-----------+-------+---------+-----------+

 

 

 

Messages

--------

 

+----------------------+------------+

|message id            |occurrences |

+======================+============+

|bad-whitespace        |2           |

+----------------------+------------+

|missing-final-newline |1           |

+----------------------+------------+

 

 

 

Global evaluation

-----------------

Your code has been rated at 2.50/10

 

其中my_math.py文件如下：

""""

A simple math module.

"""

__revision__='0.1'

def product(factor1,factor2):

   'The product of two numbers'

    return factor1 * factor2

 

当然就算有pychecker和pylint我们也需要进行手动调式，手动调试是建立在从程序的特殊情况中总结出的知识的基础上的。

2.  分析

在试图让代码提速前，有个非常重要的规则需要注意：

不成熟的优化是万恶之源。--Donald Knuth

                  此外还有KISS原则，KISS=KeepIt Small and Simple.

                  如果必须进行优化，那么绝对要在优化之前进行分析。

标准库包含了一个叫做profile的分析模块。

例如代码如下：

import profile

from my_math import product

profile.run('product(1,2)')

 

其中my_math.py文件内容如下：

""""

A simple math module.

"""

__revision__='0.1'

def product(factor1,factor2):

    'The product of two numbers'

    return factor1 *factor2

最后输出如下：

         4 function calls in 0.031 seconds

 

   Ordered by: standard name

 

   ncalls tottime  percall  cumtime percall filename:lineno(function)

        1   0.031    0.031    0.031   0.031 :0(setprofile)

        1   0.000    0.000    0.000   0.000 <string>:1(<module>)

        1   0.000    0.000    0.000   0.000 my_math.py:5(product)

        1   0.000    0.000    0.031   0.031 profile:0(product(1,2))

        0   0.000             0.000          profile:0(profiler)