通过xsd schema结构来验证xml是否合法

 import sys
 import StringIO
 import lxml

 from lxml import etree
 from StringIO import StringIO

 # Construct XML relevant to the XML schema we're validating against. By altering the string, adding/removing elements
 # we can force different errors to occur when validating.
 xml = StringIO('''
 <CompanyDataRequest xmlns="http://xmlgw.companieshouse.gov.uk" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://xmlgw.companieshouse.gov.uk http://xmlgw.companieshouse.gov.uk/v2-1/schema/CompanyData-v2-2.xsd">
     <CompanyNumber>06937730</CompanyNumber>
     <CompanyAuthenticationCode>123456</CompanyAuthenticationCode>
     <MadeUpDate>2010-06-30x</MadeUpDate>
 </CompanyDataRequest>
 ''')

 # Clear any previous errors
 lxml.etree.clear_error_log()

 try:
     # Get the XML schema to validate against
     schema = lxml.etree.XMLSchema(file = 'http://xmlgw.companieshouse.gov.uk/v2-1/schema/CompanyData-v2-2.xsd')
     # Parse string of XML
     xml_doc = lxml.etree.parse(xml)
     # Validate parsed XML against schema returning a readable message on failure
     schema.assertValid(xml_doc)
     # Validate parsed XML against schema returning boolean value indicating success/failure
     print 'schema.validate() returns "%s".' % schema.validate(xml_doc)

 except lxml.etree.XMLSchemaParseError, xspe:
     # Something wrong with the schema (getting from URL/parsing)
     print "XMLSchemaParseError occurred!"
     print xspe

 except lxml.etree.XMLSyntaxError, xse:
     # XML not well formed
     print "XMLSyntaxError occurred!"
     print xse

 except lxml.etree.DocumentInvalid, di:
     # XML failed to validate against schema
     print "DocumentInvalid occurred!"

     error = schema.error_log.last_error
     if error:
         # All the error properties (from libxml2) describing what went wrong
         print 'domain_name: ' + error.domain_name
         print 'domain: ' + str(error.domain)
         print 'filename: ' + error.filename # '<string>' cos var is a string of xml
         print 'level: ' + str(error.level)
         print 'level_name: ' + error.level_name # an integer
         print 'line: ' + str(error.line) # a unicode string that identifies the line where the error occurred.
         print 'message: ' + error.message # a unicode string that lists the message.
         print 'type: ' + str(error.type) # an integer
         print 'type_name: ' + error.type_name

封装类

 #!/usr/bin/env python
 # -*- coding:utf-8 -*-
 # Author:Eric.yue

 import os
 import lxml.etree as ET
 from StringIO import StringIO
 import chardet

 class R3xmlCheck(object):
     def __init__(self, element_xml):
         self.elem_xml = element_xml

     def validate_xsd_xml(self, f_xml, elem_xsd):
         try:
             elem_xsd = elem_xsd.encode('utf-8')
             xsd_doc = StringIO(elem_xsd)
             xml_doc = StringIO(f_xml)
             xmlschema_doc = ET.parse(xsd_doc)
             xmlschema = ET.XMLSchema(xmlschema_doc)
             xml = ET.parse(xml_doc)
             xmlschema.assertValid(xml)
             print 'schema.validate() returns "%s".' % xmlschema.validate(xml)

         except ET.XMLSchemaParseError, xspe:
             # Something wrong with the schema (getting from URL/parsing)
             print "XMLSchemaParseError occurred!"
             print xspe

         except ET.XMLSyntaxError, xse:
             # XML not well formed
             print "XMLSyntaxError occurred!"
             print xse

         except ET.DocumentInvalid, di:
             # XML failed to validate against schema
             print "DocumentInvalid occurred!"

             error = xmlschema.error_log.last_error
             if error:
                 # All the error properties (from libxml2) describing what went wrong
                 print 'domain_name: ' + error.domain_name
                 print 'domain: ' + str(error.domain)
                 print 'filename: ' + error.filename  # '<string>' cos var is a string of xml
                 print 'level: ' + str(error.level)
                 print 'level_name: ' + error.level_name  # an integer
                 print 'line: ' + str(error.line)  # a unicode string that identifies the line where the error occurred.
                 print 'message: ' + error.message  # a unicode string that lists the message.
                 print 'type: ' + str(error.type)  # an integer
                 print 'type_name: ' + error.type_name

     def run(self):
         res = self.validate_xml(self.elem_xml)
         if res["result"] is not True:
             return res["info"]

         elem_xsd = self.get_xsd()

         with open(self.elem_xml) as f:
             f_xml = f.read()
             chardet_info = chardet.detect(f_xml)
             if chardet_info['encoding'] == 'ascii':
                 f_xml = f_xml.encode('utf-8')
             self.validate_xsd_xml(f_xml.strip(),elem_xsd)

     # matching schemaLocation url
     def get_xsd(self):
         with open("./xsd/multicacheschemas/MCCI_IN200100UV01.xsd") as f:
             elem_xsd = f.read()
             return elem_xsd

     def validate_xml(self, exml):
         rinfo = {}
         if os.path.exists(exml):
             try:
                 ET.parse(exml)
                 rinfo['result'] = True
             except Exception as err:
                 rinfo['result'] = False
                 rinfo['info'] = 'Parsing error info:{0}'.format(err)
         return rinfo

 if __name__ == "__main__":
     aa = R3xmlCheck("./xsd/aa.xml")
     aa.run()