代碼的持續改進

#!/usr/bin/env python# -*- coding: utf-8 -*-#原廠模式def createDomTree(htmlStream,type='soup'):    if type == "soup":        return tnDomTreeWithSoup(htmlStream)        if type == "lxml":        return tnDomTreeWithlXml(htmlStream)        return None    #外部依賴的介面，它屏蔽了我具體使用的第三方庫class tnDomTree:    #public    def __init__(self,htmlStream):        self.htmlStream = htmlStream        def getLinkList(self):        return self._getElementByTagName('a')#提取子類共有行為到基類        def getImageList(self):        return self._getElementByTagName('img')#提取子類共有行為到基類        def elementToString(self,element):        pass        def getAttrValueOfElement(self,element,attName):        pass        #private    def _getElementByTagName(self,tagName):#子類實現該方法        pass#使用BeautifulSoup的類from BeautifulSoup import BeautifulSoupclass tnDomTreeWithSoup(tnDomTree):    def __init__(self,htmlStream):        tnDomTree.__init__(self,htmlStream)        self._tree = BeautifulSoup(self.htmlStream)            def _getElementByTagName(self,tagName):        return self._tree.findAll(tagName)        def elementToString(self,element):        return str(element)        def getAttrValueOfElement(self,element,attName):        if hasattr(element,attName):            return str(element[attName])        else:            return ""#使用lxmlfrom lxml.html import tostringfrom lxml.html.soupparser import fromstringclass tnDomTreeWithlXml(tnDomTree):    def __init__(self,htmlStream):        tnDomTree.__init__(self,htmlStream)        self._tree = fromstring(self.htmlStream)            def _getElementByTagName(self,tagName):        list = []        for i in self._tree.iter():            if i.tag == tagName:                list.append(i)        return list        def elementToString(self,element):        return tostring(element)        def getAttrValueOfElement(self,element,attName):        if attName in set(element.keys()):            return str(element.attrib[attName])        else:            return ""if __name__ == "__main__":    s1 = '''    <p>BEIAI</p> <p><img src="/attachment/28" alt=""></p>    '''    print s1    domtree = createDomTree(s1,'soup')    list = domtree.getImageList()    for i in list:        print domtree.getAttrValueOfElement(i,'src')        print domtree.elementToString(i)

為什麼能抽出一個公用的介面出來？本質上是因為這些東西有相似性，而具體到代碼的設計，就應該設計成和其他類相似的API，否則類的設計就是不對的