AnyFilter 0.05

动机

导致这种解决方案的问题是需要从 最终用户。通常，尤其是当用户是客户机时，这些数据可以 不依赖于满足系统的输入规范。

这通常会导致这些次优解之一：

• 使用自定义脚本预处理每个客户端的数据
• 向核心产品添加一堆if语句或其他类似逻辑
• 试图使转换脚本具有足够的通用性以供重用，因此 使它们在主要用途上不那么有用，也更难调试
• 将数据硬编码为转换脚本

通常每个客户机或项目都有一个自定义转换， 因此，这些解决方案的可伸缩性不好。

anyfilter的目标是：

• 最小化主代码库中的自定义代码量
• 以可移植格式将配置存储在应用程序数据库之外
• 允许在不部署代码的情况下更新配置数据（特权用户 甚至可以通过某种gui编辑配置）

优点

• 创建可插入的数据筛选器
• 将配置存储在应用程序数据库之外
• 轻松备份和恢复配置
• 跨服务器轻松复制配置

计划功能

• 导出和导入配置
• 将配置转换为HTML表单和从HTML表单转换为易于使用的前端功能
• 轻松恢复到以前的配置
• 综合单元测试

示例用法

#!/usr/bin/env python"""
This is a simple example of the use of the Filter class. In this case, a
dictionary has some keys renamed. This is a trivial example; filters
can be as complex as required.
"""fromanyfilterimportFilterclassNameFilter(Filter):"""
    A filter that changes the names of dictionary keys.

    'data' should be an iterable of dictionaries
    """def__call__(self,data):"""
        The contents of this function are the least-important part of
        this demo. This is where you custom code will go, doing whatever
        it is you need with whatever config format and content you need.
        """forrecindata:# The config for this filter is a dictionary where the# key is the key name to replace, and the value is the new name.# update values in "data" dictforkey,valueinself.config.items():ifkeyinrec:rec[value]=rec[key]delrec[key]returndataif__name__=='__main__':importos# for dealing with the environment variable manually# set environment variable for demo purposesoriginal_envvar=os.environ.get('FILTER_CONFIG_DIR','')os.environ['FILTER_CONFIG_DIR']='/tmp'# Instantiate subclass. The only argument is the uid of the subject# of the filter. For example, if you need to store different rules# per user of your site, you might use the user's primary key here.# This allows storage of configs per filter *and* per user.name_filter=NameFilter('foo')# Set some filter items. This normally won't be a part of the flow.# It's here for demo purposes. In normal usage, the config would# already be set and probably rarely updated.name_filter.config={'dog':'canine','cat':'feline','horse':'equine',}name_filter.save_config(user='example')data=[{'cat':'meow','dog':'woof','horse':'neigh','foo':'bar',}]printdata# originalprintname_filter(data)# altered# Put it back like we found it, just to be good citizens.os.environ['FILTER_CONFIG_DIR']=original_envvar

样本输出

[{'horse': 'neigh', 'foo': 'bar', 'dog': 'woof', 'cat': 'meow'}]
[{'equine': 'neigh', 'feline': 'meow', 'canine': 'woof', 'foo': 'bar'}]