python自訂解析簡單xml格式檔案

因為公司內部的介面返回的字串支援2種形式：php數組，xml；結果php數組python不能直接用，而xml字串的格式不是標準的，所以也不能用標準模組解析。【不標準的地方是某些節點會的名稱是以數字開頭的】，所以寫個簡單的腳步來解析一下檔案，用來做介面測試。

#!/usr/bin/env python  #encoding: utf-8import reclass xmlparse:        def __init__(self, xmlstr):        self.xmlstr = xmlstr        self.xmldom = self.__convet2utf8()        self.xmlnodelist = []        self.xpath = ''            def __convet2utf8(self):        headstr = self.__get_head()        xmldomstr = self.xmlstr.replace(headstr, '')        if 'gbk' in headstr:            xmldomstr = xmldomstr.decode('gbk').encode('utf-8')        elif 'gb2312' in headstr:            xmldomstr = self.xmlstr.decode('gb2312').encode('utf-8')        return xmldomstr        def __get_head(self):        headpat = r'<\?xml.*\?>'        headpatobj = re.compile(headpat)        headregobj = headpatobj.match(self.xmlstr)          if headregobj:            headstr = headregobj.group()            return headstr        else:            return ''                def parse(self, xpath):         self.xpath = xpath        xpatlist = []        xpatharr = self.xpath.split('/')        for xnode in xpatharr:            if xnode:                spcindex = xnode.find('[')                if spcindex > -1:                    index = int(xnode[spcindex+1:-1])                    xnode = xnode[:spcindex]                            else:                    index = 0;                temppat = ('<%s>(.*?)</%s>' % (xnode, xnode),index)                               xpatlist.append(temppat)                xmlnodestr = self.xmldom        for xpat,index in xpatlist:            xmlnodelist = re.findall(xpat,xmlnodestr)             xmlnodestr = xmlnodelist[index]            if xmlnodestr.startswith(r'<![CDATA['):                xmlnodestr = xmlnodestr.replace(r'<![CDATA[','')[:-3]                        self.xmlnodelist = xmlnodelist        return xmlnodestr                                        if '__main__' == __name__:    xmlstr = '<?xml version="1.0" encoding="utf-8" standalone="yes" ?><resultObject><a><product_id>aaaaa</product_id><product_name><![CDATA[bbbbb]]></a><b><product_id>bbbbb</product_id><product_name><![CDATA[bbbbb]]></b></product_name></resultObject>'                    xpath1 = '/product_id'    xpath2 = '/product_id[1]'    xpath3 = '/a/product_id'    xp = xmlparse(xmlstr)      print 'xmlstr:',xp.xmlstr    print 'xmldom:',xp.xmldom      print '------------------------------'    getstr = xp.parse(xpath1)    print 'xpath:',xp.xpath          print 'get list:',xp.xmlnodelist         print 'get string:', getstr                     print '------------------------------'    getstr = xp.parse(xpath2)    print 'xpath:',xp.xpath          print 'get list:',xp.xmlnodelist         print 'get string:', getstr             print '------------------------------'    getstr = xp.parse(xpath3)    print 'xpath:',xp.xpath          print 'get list:',xp.xmlnodelist         print 'get string:', getstr                    

運行結果：

xmlstr: <?xml version="1.0" encoding="utf-8" standalone="yes" ?><resultObject><a><product_id>aaaaa</product_id><product_name><![CDATA[bbbbb]]></a><b><product_id>bbbbb</product_id><product_name><![CDATA[bbbbb]]></b></product_name></resultObject>xmldom: <resultObject><a><product_id>aaaaa</product_id><product_name><![CDATA[bbbbb]]></a><b><product_id>bbbbb</product_id><product_name><![CDATA[bbbbb]]></b></product_name></resultObject>------------------------------xpath: /product_idget list: ['aaaaa', 'bbbbb']get string: aaaaa------------------------------xpath: /product_id[1]get list: ['aaaaa', 'bbbbb']get string: bbbbb------------------------------xpath: /a/product_idget list: ['aaaaa']get string: aaaaa

因為返回的xml格式比較簡單，沒有帶屬性的節點，所以處理起來就比較簡單了。但測試還是發現有一個bug。即當相同節點嵌套時會出現正則匹配出問題，該問題的可以通過避免在xpath中出現有嵌套節點的名稱來解決，否則只有重寫複雜的機制了。