在Python中实现PageFactory模式

　　关于 PageFactory 的概念主要是Java中内置了PageFactory类。

import org.openqa.selenium.support.PageFactory;  

……

例子，http://libin0019.iteye.com/blog/1260090   

　　Python（Selenium）中没有这个类。 PageFactory 的概念和Page Object应该类似，属于一种设计模式。所以并不局限于语言及场景。于是，好奇，既然Java有，那Python也应该有类似的玩法。还真让我给找到了类似的实现。

原文在此：https://jeremykao.wordpress.com/2015/06/10/pagefactory-pattern-in-python/

于是，就借助谷歌翻译加代码运行，弄懂了这哥们想要利用PageFactory 模式实现个什么东西，为了便于你的理解，我这里搬运过来给下结论。

selenium in python中的元素定位是这样的：

find_element_by_id("kw")
find_element_by_xpath("//*[@id='kw']")

或者是这样的：

from selenium.webdriver.common.by import By

find_element(By.ID,"kw")
find_element(By.XPATH,"//*[@id='kw']")

通过PageFactory 模式的实现可以把元素定位变成这样的：

from pageobject_support import callable_find_by as find_by

find_by(id_="kw")
find_by(xpath="//*[@id='kw']")

 别看小小的改动，它其实使代码更容易的阅读和理解。

核心实现就是pageobject_support.py文件：

__all__ = ['cacheable', 'callable_find_by', 'property_find_by']

def cacheable_decorator(lookup):
    def func(self):
        if not hasattr(self, '_elements_cache'):
            self._elements_cache = {} # {callable_id: element(s)}
        cache = self._elements_cache

        key = id(lookup)
        if key not in cache:
            cache[key] = lookup(self)
        return cache[key]

    return func

cacheable = cacheable_decorator

_strategy_kwargs = ['id_', 'xpath', 'link_text', 'partial_link_text',
                    'name', 'tag_name', 'class_name', 'css_selector']

def _callable_find_by(how, using, multiple, cacheable, context, driver_attr, **kwargs):
    def func(self):
        # context - driver or a certain element
        if context:
           ctx = context() if callable(context) else context.__get__(self) # or property
        else:
           ctx = getattr(self, driver_attr)

        # 'how' AND 'using' take precedence over keyword arguments
        if how and using:
            lookup = ctx.find_elements if multiple else ctx.find_element
            return lookup(how, using)

        if len(kwargs) != 1 or kwargs.keys()[0] not in _strategy_kwargs :
            raise ValueError(
                "If 'how' AND 'using' are not specified, one and only one of the following "
                "valid keyword arguments should be provided: %s." % _strategy_kwargs)

        key = kwargs.keys()[0]; value = kwargs[key]
        suffix = key[:-1] if key.endswith('_') else key # find_element(s)_by_xxx
        prefix = 'find_elements_by' if multiple else 'find_element_by'
        lookup = getattr(ctx, '%s_%s' % (prefix, suffix))
        return lookup(value)

    return cacheable_decorator(func) if cacheable else func

def callable_find_by(how=None, using=None, multiple=False, cacheable=False, context=None, driver_attr='_driver', **kwargs):
    return _callable_find_by(how, using, multiple, cacheable, context, driver_attr, **kwargs)

def property_find_by(how=None, using=None, multiple=False, cacheable=False, context=None, driver_attr='_driver', **kwargs):
    return property(_callable_find_by(how, using, multiple, cacheable, context, driver_attr, **kwargs))

然后，我再帖一下具体的例子：

from pageobject_support import callable_find_by as find_by
from selenium import webdriver

class BaiduSearchPage(object):

    def __init__(self, driver):
        self._driver = driver

    search_box = find_by(id_="kw")
    search_button = find_by(id_='su')

    def search(self, keywords):
        self.search_box().clear()
        self.search_box().send_keys(keywords)
        self.search_button().click()

if __name__ == '__main__':
    driver = webdriver.Chrome()
    driver.get("https://www.baidu.com")
    BaiduSearchPage(driver).search("selenium")
    driver.close()

同样封装了8种定位方法：

• id_ (为避免与内置的关键字ID冲突，所以多了个下划线的后缀)
• name
• class_name
• css_selector
• tag_name
• xpath
• link_text
• partial_link_text

当然，这只是PageFactory 模式的一种表现形式而已。除此之外，我还找到了另外一个PageFactory模式的例子。

https://github.com/mattfair/SeleniumFactory-for-Python

这哥们是利用PageFactory模式把驱动的创建做了封装，感兴趣可以了解一下。

搬运完了，准备过年。新年快了~！！！