记一次项目中协议解析代码重构

现在项目里的协议解析很混乱&＃xff0c;而且不能复用父类的解析代码&＃xff0c;新人扩展协议很麻烦&＃xff0c;尤其会把代码写的很丑陋&＃xff0c;比如下面这样&＃xff1a;

要解析的协议格式&＃xff1a;

解析代码&＃xff0c;每个子类都得实现自己的&＃xff1a; def parse(data, encoded&＃61;True):

syncUnitDOM &＃61; parseXmlString(data)syncUnitNode &＃61; syncUnitDOM.childNodes[0]paramMap &＃61; {}efbList &＃61; []depList &＃61; []for syncUnitChildNode in syncUnitNode.childNodes:if syncUnitChildNode.nodeName &＃61;&＃61; PARAMETER_TAG:name &＃61; str(syncUnitChildNode.getAttribute(NAME_TAG))value &＃61; str(syncUnitChildNode.getAttribute(VALUE_TAG))paramMap[name] &＃61; valueelif syncUnitChildNode.nodeName &＃61;&＃61; EFBLIST_TAG:efbListParameterList &＃61; syncUnitChildNode.getElementsByTagName(PARAMETER_TAG)for efbListNode in efbListParameterList:if str(efbListNode.getAttribute(NAME_TAG)) &＃61;&＃61; EFB_NAME:efbList.append(str(efbListNode.getAttribute(VALUE_TAG)))elif syncUnitChildNode.nodeName &＃61;&＃61; DEPLIST_TAG:for contentNode in syncUnitChildNode.childNodes:if contentNode.nodeName &＃61;&＃61; CONTENT_TAG:content &＃61; {}for conParamNode in contentNode.getElementsByTagName(PARAMETER_TAG):paramName &＃61; str(conParamNode.getAttribute(NAME_TAG))if paramName in ENCODED_TAGS and encoded:content[paramName] &＃61; decode(str(conParamNode.getAttribute(VALUE_TAG)))else:content[paramName] &＃61; str(conParamNode.getAttribute(VALUE_TAG))depList.append(content)


不管别人舒不舒服&＃xff0c;我受不了了。这种情况必须得改善&＃xff0c;一开始想到的是go 语言解析xml那种方式&＃xff0c;后来又觉得像声明orm model那样更方便&＃xff0c;看了peewee这个orm框架后就开工了&＃xff0c;下面是一个model的描述&＃xff1a;


class SyncUploadRequestUnit(Model):content_name &＃61; ParameterField(&＃39;ContentName&＃39;)data_type &＃61; ParameterField(&＃39;DataType&＃39;)publish_date &＃61; ParameterField(&＃39;PublishDate&＃39;)effective_time &＃61; ParameterField(&＃39;EffectiveTime&＃39;)effective_date &＃61; ParameterField(&＃39;EffectiveDate&＃39;)filename &＃61; ParameterField(&＃39;FileName&＃39;)unassign_flag &＃61; ParameterField(&＃39;UnassignFlag&＃39;)crc_value &＃61; ParameterField(&＃39;CRCValue&＃39;)assignment_id &＃61; ParameterField(&＃39;AssignmentId&＃39;)title &＃61; ParameterField(&＃39;Title&＃39;)keywords &＃61; ParameterField(&＃39;Keywords&＃39;)version &＃61; ParameterField(&＃39;Version&＃39;)efb_list &＃61; StringListField(&＃39;EfbList&＃39;, &＃39;EfbName&＃39;)dep_list &＃61; ModelListField(&＃39;DepList&＃39;, Content)


看着很直观是不是&＃xff0c;下面是其基类


import base64
import xml.etree.ElementTree as ETBEGIN_TAG &＃61; &＃39;CM&＃39;
PARAMETER_TAG &＃61; &＃39;Parameter&＃39;
ENCODED_TAGS &＃61; [&＃39;DataType&＃39;, &＃39;FileName&＃39;, &＃39;ContentName&＃39;, &＃39;Revision&＃39;, &＃39;Title&＃39;]def encode(value):return base64.encodestring(value.encode(&＃39;utf-8&＃39;))def decode(value):return base64.decodestring(value).decode(&＃39;utf-8&＃39;)class FieldDescriptor(object):def __init__(self, name, field):self.name &＃61; nameself.field &＃61; fielddef __get__(self, instance, cls):value &＃61; instance._data.get(self.field.tag, None)return value def __set__(self, instance, value):instance._data[self.field.tag] &＃61; valueclass TagField(object):def __init__(self, tag):self.tag &＃61; tag\def parse(self, root, setter, encoded&＃61;False):for element in root.iterfind(self.tag):setter(self.tag, element.text, encoded)def transfer_to_xml(self, parent, getter, encoded&＃61;False):element &＃61; ET.SubElement(parent, self.tag)element.text &＃61; getter(self.tag, encoded) or &＃39;&＃39;class ParameterField(object):def __init__(self, name):self.tag &＃61; namedef parse(self, root, setter, encoded&＃61;False):for element in root.iterfind(PARAMETER_TAG):setter(element.get(&＃39;name&＃39;), element.get(&＃39;value&＃39;), encoded)def transfer_to_xml(self, parent, getter, encoded&＃61;False):attrs &＃61; {&＃39;name&＃39;: self.tag, &＃39;value&＃39;: getter(self.tag, encoded)}ET.SubElement(parent, PARAMETER_TAG, attrs)class StringListField(object):def __init__(self, container_name, element_name):self.tag &＃61; container_nameself.element_name &＃61; element_namedef parse(self, root, setter, encoded&＃61;False):children &＃61; []setter(self.tag, children)container &＃61; root.find(self.tag)if container is None:returnfor element in container.iterfind(PARAMETER_TAG):children.append(element.get(&＃39;value&＃39;))def transfer_to_xml(self, parent, getter, encoded&＃61;False):value_list &＃61; getter(self.tag, encoded)if value_list is None or len(value_list) &＃61;&＃61; 0:returnparent &＃61; ET.SubElement(parent, self.tag)for value in value_list:attrs &＃61; {&＃39;name&＃39;: self.element_name, &＃39;value&＃39;: value}ET.SubElement(parent, PARAMETER_TAG, attrs)class ModelListField(object):def __init__(self, container_name, model):self.tag &＃61; container_nameself.model &＃61; modelself.model_tag &＃61; model.__name__def parse(self, root, setter, encoded&＃61;False):children &＃61; []setter(self.tag, children)container &＃61; root.find(self.tag)if container is None:returnfor element in container.iterfind(self.model_tag):child &＃61; self.model()child.parse_element(element, encoded)children.append(child)def transfer_to_xml(self, parent, getter, encoded&＃61;False):value_list &＃61; getter(self.tag, encoded)if value_list is None or len(value_list) &＃61;&＃61; 0:returnparent &＃61; ET.SubElement(parent, self.tag)for value in value_list:value.as_xml(parent, encoded)class ModelField(object):def __init__(self, model):self.model &＃61; modelself.tag &＃61; model.__name__def parse(self, root, setter, encoded&＃61;False):element &＃61; root.find(self.tag)if element is not None:child &＃61; self.model()child.parse_element(element, encoded)setter(self.tag, child)def transfer_to_xml(self, parent, getter, encoded&＃61;False):value &＃61; getter(self.tag, encoded)if value is not None:value.as_xml(parent, encoded)FIELD_LIST &＃61; [TagField, ParameterField, StringListField, ModelField, ModelListField]class BaseModel(type):def __new__(cls, name, bases, attrs):if not bases:return super(BaseModel, cls).__new__(cls, name, bases, attrs)fields &＃61; {TagField: [],ParameterField: [],ModelField: [],StringListField: [],ModelListField: [],}for attr, field in attrs.items():field_class &＃61; type(field)if field_class in fields:fields[field_class].append(field)attrs[attr] &＃61; FieldDescriptor(attr, field)attrs[&＃39;_fields&＃39;] &＃61; fieldsreturn super(BaseModel, cls).__new__(cls, name, bases, attrs)class Model(object):__metaclass__ &＃61; BaseModeldef __init__(self, data&＃61;None):self._data &＃61; data or {}def parse(self, data, encoded&＃61;False):return self.parsestring(data, encoded)def parsestring(self, data, encoded&＃61;False):root &＃61; ET.fromstring(data)return self.parse_element(root, encoded)def parse_element(self, root, encoded&＃61;False):for field_list in self._fields.values():for field in field_list:field.parse(root, self.__set, encoded)return selfdef as_xml(self, parent&＃61;None, encoded&＃61;False):tag &＃61; self.__class__.__name__if parent is not None:parent &＃61; ET.SubElement(parent, tag)else:parent &＃61; ET.Element(BEGIN_TAG)for field_class in FIELD_LIST:for field in self._fields.get(field_class, []):field.transfer_to_xml(parent, self.__get, encoded)return parentdef as_pretty_xml(self, encoded&＃61;False):tree &＃61; self.as_xml(encoded&＃61;encoded)self.indent(tree)return ET.tostring(tree)def indent(self, elem, level&＃61;0):i &＃61; "\n" &＃43; level * " "if len(elem):if not elem.text or not elem.text.strip():elem.text &＃61; i &＃43; " "if not elem.tail or not elem.tail.strip():elem.tail &＃61; ifor elem in elem:self.indent(elem, level&＃43;1)if not elem.tail or not elem.tail.strip():elem.tail &＃61; ielse:if level and (not elem.tail or not elem.tail.strip()):elem.tail &＃61; idef __set(self, name, value, encoded&＃61;False):if encoded and name in ENCODED_TAGS:value &＃61; decode(value)self._data[name] &＃61; valuedef __get(self, name, encoded&＃61;False):value &＃61; self._data.get(name, None)if encoded and value and name in ENCODED_TAGS:value &＃61; encoded(value)return value