FangZi

Quickly Start

Console

python setup.py install

Python Console

from fangzi import FangZi
handler = FangZi()

Or use it like a script

python fangzi.py --status

---

What is FangZi

FangZi is a simple tool for manage functions in MongoDB.

Dynamic Function Check

Consider this case:

# A data need many check
if rule1_check(data) and rule2_check(data1, data2) and ...:
    # Yes it pass all the rule we made
    # Then we can do the formal process
    ...

If we have something need so many rule-check in our project, and those functions of rules many be changed or updated frequently.

It is hard to manage then.

Now here is another way:

flag = True
for func in all_funcs:
    try:
        exec func.code
    except Exception, result:
        flag = result
        
if flag:
    # It pass all the rule we made and let's talk bussiness

See, we make the check part invariant.

If we want to manage the functions, all we need is to handle the content of func.

So that we get a dynamic-function-check solution.

How it works?

The point is that we store all of our check-functions into MongoDB, each function take one Ducument.

We don't need the functions' head and its return, we parse them and make it into the code we need.

Source Function

def check(*args, **kwargs):
    # ... do something
    if valid:
        return True
    else:
        return False

Parse Code We Need

# ... do something
if valid:
    result = True; raise Exception(result)
else:
    result = False; raise Exception(result)

This code can be execute by exec and the result will be catch by except.

So we get a full solution for the dynamic-function-check.

Usage

Install MongoDB: pip install -r requirement.txt

Make sure your local MongoDB server is running

Config the settings in settings.py

Parse from files and launch into database: python fangzi.py -p -l

Show the functions' status in database: python fangzi.py -s

Wake up the functions by Flag/Group/Name: python fangzi.py --wake --flag CODE

More

A full example is provided in example/

Hope you come to discuss about the solution and any criticism or suggestion is welcome:)

中文介绍

什么是房子

房子(FangZi)是一个简单的脚本工具，确切的说是一个工具类，可以在你的代码中直接import嵌入使用。

动态规则检查

看看这个例子：

# A data need many check
if rule1_check(data) and rule2_check(data1, data2) and ...:
    # Yes it pass all the rule we made
    # Then we can do the formal process
    ...

当某些数据需要接受大量函数检测时（如合法性检测、特征值封禁、日志分类统计、流水线加工），我们要花费大量的代价去维护一系列函数。

特别当某些检查特征的函数需要经常变动时，当对某些函数进行更改或者添加删除时，这种方式的工作量很大，不易于管理。

现在看看另一个例子：

flag = True
for func in all_funcs:
    try:
        exec func.code
    except Exception, result:
        flag = result
        
if flag:
    # It pass all the rule we made and let's talk bussiness

我们把所有的函数一并放在了all_funcs里，这样把检测部分的代码固定住了，当我们对函数进行管理时，只需要对func.code进行管理就可以了。

因此我们得到了一个解决方案：

将所有的函数储存在数据库中，并在数据库中通过数据库读写来进行动态管理

当需要使用时，从数据库中取出来，使用exec执行其代码部分

怎么做

房子(FangZi)是用来把普通的包含函数的源文件转化并写入数据库中的工具，并且提供了管理的方法。

我们使用MongoDB进行储存，一个函数(function)对应一个数据库中的文档(Document)

那么怎么用呢？

源函数

def check(*args, **kwargs):
    # ... do something
    if valid:
        return True
    else:
        return False

转化后的函数

# ... do something
if valid:
    result = True; raise Exception(result)
else:
    result = False; raise Exception(result)

我们将函数中所有的return Bool替换为了result = Bool; raise Exception(result)，并在使用的时候用try-except和exec去执行+捕捉。

这样就获得了函数执行的结果。

使用

见上文 Usage