python 脚本 通过解析 DDL 更新PDM
秦俊友
需求：
上线工作涉及到表结构更新时，由开发人员提交变更的sql语句由DBA操作，为了保障表结构变更与PDM同步并且自动化，写了python脚本，解析DDL语句更新PDM文件。
PDM是XML文件为脚本更新提供可能。
代码如下：
PDM解析操作脚本：
#!/usr/bin/python
# -*- coding: utf-8 -*- #
from xml.dom.minidom import parse
import xml.dom.minidom
class PDMHandler(object):
TBL_ATTR_LIST = ["Code"]
COL_ATTR_LIST = ["Code"]
tableNodes = {}
columnNodes = {}
modelTablesNode = None
ph = None
filename = 'package.pdm'
def __init__(self):
self.init();
return
def init(self):
self.ph = PDMHandler.parse(self.filename)
for pkg in PDMHandler.getPkgNodes(self.ph):
for tbl in PDMHandler.getTblNodesInPkg(pkg):
tbl_attrs = PDMHandler.getTblAttrs(tbl)
self.tableNodes[tbl_attrs['Code']] = tbl
for col in PDMHandler.getColNodesInTbl(tbl):
col_attrs = PDMHandler.getColAttrs(col)
self.columnNodes[tbl_attrs['Code'] + '_' + col_attrs['Code']] = col
@staticmethod
def parse(pdmfilename):
"""
@brief PDM文件导入函数
@param[in] pdmfilename pdm文件名
@return pdm文件的DOM结构
"""
return xml.dom.minidom.parse(pdmfilename)
@staticmethod
def __get_nodes_by_path(parent, xml_path):
"""
@brief 按路径取节点核心函数,通过传入一个节点和一个路径
(类文件系统路径)的字符串,获得相应的子结点(或集合).
例1: __get_nodes_by_path(node,"a/b/c|3/d")
即获得node下a下b下第3个c节点下的所有d节点的一个
list.
例2: __get_nodes_by_path(node,"a/b/c|3/d|2")
即获得node下a下b下第3个c节点下的第2个d节点.
注意:此处返回的不是list,而是节点
@param[in] parent DOM中的父节点
@return 节点list或节点句柄
"""
curr_node = parent
for tag in xml_path.split("/")[0:-1]:
tag_desc = tag.split("|")
tag_name, tag_index = tag_desc[0], (int(tag_desc[1]) if len(tag_desc) == 2 else 0)
child_nodes = []
for child_node in curr_node.childNodes:
if child_node.nodeName == tag_name:
child_nodes.append(child_node)
if len(child_nodes) < tag_index + 1:
return []
curr_node = child_nodes[tag_index]
# -- 最后一级路径特殊处理 -- #
tag = xml_path.split("/")[-1]
tag_desc = tag.split("|")
tag_name, tag_index = tag_desc[0], (int(tag_desc[1]) if len(tag_desc) == 2 else None)
child_nodes = []
for child_node in curr_node.childNodes:
if child_node.nodeName == tag_name:
child_nodes.append(child_node)
if tag_index == None:
return child_nodes
elif len(child_nodes) < tag_index + 1:
return []
else:
curr_node = child_nodes[tag_index]
return curr_node
@staticmethod
def __get_attrs_by_list(parent, attr_list):
"""
@brief 取PDM属性(注意:此属性非xml属性）
背景:PDM文件中主要节点(o:Table,o:Column...)的具体属性
一般以子节点形式表示，而不以通常的XML属性表示.此
函数传入一个attr_list,输出一个字典表示的:
{ "attr":"value" ...}
@param[in] parent 需要被取属性的节点句柄
@param[in] attr_list 一个字符串的list["attr1","attr2"...],代表要取的属性名集合
@return 返回一个dict:{"attr1":"value" ...}
"""
ret_dict = {}
for attr in attr_list:
ret_dict[attr] = ""
for child in parent.childNodes:
if child.nodeName == "a:" + attr:
ret_dict[attr] = child.childNodes[0].data
break
return ret_dict
@staticmethod
def __get_pkgnodes_recursively(o_pkg):
# -- 需要传入一个o:Model/o:Package节点 --#
if o_pkg.nodeName != "o:Model" and o_pkg.nodeName != "o:Package":
return []
ret_list = []
subpkgs = PDMHandler.__get_nodes_by_path(o_pkg, "c:Packages/o:Package")
if subpkgs != None:
for subpkg in subpkgs:
ret_list.append(subpkg)
ret_list = ret_list + PDMHandler.__get_pkgnodes_recursively(subpkg)
else:
return []
return ret_list
@staticmethod
def getPkgNodes(hpdm):
"""
@brief 获取pdm文件中所有的o:package的句柄list
@param[in] hpdm 待处理的pdm文件DOM句柄(可通过PDMHandler.parse取得)
@return 返回pacakge的节点list:[pkg1,pkg2...],元素为DOM节点类型
"""
ret_list = []
try:
o_mdl = PDMHandler.__get_nodes_by_path(hpdm, "Model/o:RootObject/c:Children/o:Model")[0]
ret_list.append(o_mdl)
except IndexError:
print("ERROR:不是一个合法的pdm文件!")
return []
ret_list = ret_list + PDMHandler.__get_pkgnodes_recursively(o_mdl)
return ret_list
@staticmethod
def getModelTablesNodes(hpdm):
"""
@brief 获取pdm文件中所有的o:package的句柄list
@param[in] hpdm 待处理的pdm文件DOM句柄(可通过PDMHandler.parse取得)
@return 返回pacakge的节点list:[pkg1,pkg2...],元素为DOM节点类型
"""
ret_list = []
try:
o_mdl = PDMHandler.__get_nodes_by_path(hpdm, "Model/o:RootObject/c:Children/o:Model/c:Tables")
if len(o_mdl) == 0:
o_mdl = PDMHandler.__get_nodes_by_path(hpdm, "Model/o:RootObject/c:Children/o:Model")[0]
cTables = hpdm.createElement('c:Tables')
o_mdl.appendChild(cTables)
ret_list.append(cTables)
else:
ret_list.append(o_mdl[0])
except IndexError:
return []
return ret_list
@staticmethod
def getTblNodesInPkg(pkgnode):
"""
@brief 获取指定o:Package下的所有o:Table的list
@param[in] pkgnode 待处理的o:Package节点(可通过PDMHandler.getPkgNodes取得)
@return 返回o:Table的节点list:[tbl1,tbl2...],元素为DOM节点类型
"""
return PDMHandler.__get_nodes_by_path(pkgnode, "c:Tables/o:Table")
@staticmethod
def getColNodesInTbl(tblnode):
"""
@brief 获取指定o:Table下的所有o:Column的list
@param[in] tblnode 待处理的o:Table节点(可通过PDMHandler.getTblInPkg取得)
@return 返回o:Column的节点list:[col1,col2...],元素为DOM节点类型
"""
return PDMHandler.__get_nodes_by_path(tblnode, "c:Columns/o:Column")
@staticmethod
def getIdxNodesInTbl(tblnode):
"""
@brief 获取指定o:Table下的所有o:Index的list
@param[in] tblnode 待处理的o:Table节点(可通过PDMHandler.getTblInPkg取得)
@return 返回o:Index的节点list:[idx1,idx2...],元素为DOM节点类型
"""
return PDMHandler.__get_nodes_by_path(tblnode, "c:Indexes/o:Index")
@staticmethod
def getIdxColNodesInIdx(idxnode):
"""
@brief 获取指定o:Index下的所有o:IndexColumn的list
@param[in] idxnode 待处理的o:Index节点(可通过PDMHandler.getIdxNodesInTbl取得)
@return 返回o:IndexColumn的节点list:[idxcol1,idxcol2...],元素为DOM节点类型
"""
return PDMHandler.__get_nodes_by_path(idxnode, "c:IndexColumns/o:IndexColumn")
@staticmethod
def getTblAttrs(tblnode):
"""
@brief 获取指定o:Table的属性(可取的属性参见PDMHandler.TBL_ATTR_LIST)
@param[in] tblnode 待处理的o:Table节点(可通过PDMHandler.getTblNodesInPkg取得)
@return 返回一个字典dict:{"attr1":"value",...}
"""
return PDMHandler.__get_attrs_by_list(tblnode, PDMHandler.TBL_ATTR_LIST)
@staticmethod
def getColAttrs(colnode):
"""
@brief 获取指定o:Column的属性(可取的属性参见PDMHandler.COL_ATTR_LIST)
@param[in] colnode 待处理的o:Column节点(可通过PDMHandler.getColNodesInTbl取得)
@return 返回一个字典dict:{"attr1":"value",...}
"""
return PDMHandler.__get_attrs_by_list(colnode, PDMHandler.COL_ATTR_LIST)
@staticmethod
def getNodePath(node):
"""
@brief 获取指定node的路径(模拟文件系统路径表示法)
@param[in] node 待处理的节点,类型为DOM的节点类型
@return 返回一个字符串表示node的路径,形如"/foo/bar/node"
"""
curr = node
path_nodes = []
while (1):
if curr != None and curr.nodeName != "#document":
path_nodes.append(curr.tagName)
else:
break
curr = curr.parentNode
path_nodes.reverse()
path = "".join([slash + node for slash in '/' for node in path_nodes])
return path
@staticmethod
def getNode(parent, nodeName):
"""
@brief 判断结点parent的名称是否与参数名称nodeName相同，相同返回parent,否则返则None
:param parent: 节点
:param nodeName: 参数名称
:return:
"""
for node in parent.childNodes:
if node.nodeName == 'a:Name' and node.childNodes[0].data == nodeName:
return parent
return None
@staticmethod
def getTableColumnsNode(tableNode):
"""
@brief 根据表结点tableNode查询列结点
:param tableNode: 表节点
:return: 表节点下节点c:Columns
"""
for node in tableNode.childNodes:
if node.nodeName == 'c:Columns':
return node
return None
@staticmethod
def addColumn(pdm, tableCode, columnObjectList=None):
"""
@brief 增加列
:param pdm: dom文档
:return:
"""
if tableCode == None:
print('待增加列对应表编码不能为空')
return False
if columnObjectList == None or len(columnObjectList) == 0:
print('待增加列不能为空且格式为[{"code":,"name":,...}]')
return False
for columnObject in columnObjectList:
tableNode = pdm.tableNodes[tableCode]
TableColumnsNode = PDMHandler.getTableColumnsNode(tableNode)
nColumn = pdm.ph.createElement('o:Column')
TableColumnsNode.appendChild(nColumn)
nCode = pdm.ph.createElement('a:Code')
cCode = pdm.ph.createTextNode(columnObject['code'])
nCode.appendChild(cCode)
nColumn.appendChild(nCode)
pdm.columnNodes[tableCode + '_' + columnObject['code']] = nColumn
nName = pdm.ph.createElement('a:Name')
cName = pdm.ph.createTextNode(columnObject['code'] if columnObject.get('name') == None else columnObject['name'])
nName.appendChild(cName)
nColumn.appendChild(nName)
if columnObject.get('comment') != None:
nComment = pdm.ph.createElement('a:Comment')
cComment = pdm.ph.createTextNode(columnObject['comment'])
nComment.appendChild(cComment)
nColumn.appendChild(nComment)
nDataType = pdm.ph.createElement('a:DataType')
cDataType = pdm.ph.createTextNode(columnObject['dataType'])
nDataType.appendChild(cDataType)
nColumn.appendChild(nDataType)
nLength = pdm.ph.createElement('a:Length')
cLength = pdm.ph.createTextNode(columnObject['length'])
nLength.appendChild(cLength)
nColumn.appendChild(nLength)
@staticmethod
def updateColumn(pdm, tableCode, columnObjectList=None):
"""
@breif 更新列
:param pdm: dom文档
:param tableCode: 表编码
:param columnCode:
:return:
"""
if tableCode == None:
print('待修改列对应表编码不能为空')
return False
if columnObjectList == None or len(columnObjectList) == 0:
print('待增加列不能为空且格式为[{"code":,"name":},...]')
return False
for columnObject in columnObjectList:
columnNode = pdm.columnNodes[tableCode + '_' + columnObject['code']]
for key in columnObject.keys():
if key != 'code':
tf = False
for node in columnNode.childNodes:
if node.nodeName == 'a:' + key.capitalize():
node.removeChild(node.childNodes[0])
cName = pdm.ph.createTextNode(columnObject[key])
node.appendChild(cName)
tf = True
break
if tf == False:
csNode = pdm.ph.createElement('a:' + key.capitalize())
columnNode.appendChild(csNode)
cName = pdm.ph.createTextNode(columnObject[key])
csNode.appendChild(cName)
@staticmethod
def addTable(pdm, tableObject=None):
"""
@breif 增加表
:param pdm: dom文档
:return:
"""
if tableObject == None or tableObject.get('code') == None:
print('待增加表不能为空且格式为{"code":,...}')
return False
if pdm.modelTablesNode == None:
pdm.modelTablesNode = PDMHandler.getModelTablesNodes(pdm.ph)[0]
nTable = pdm.ph.createElement('o:Table')
pdm.modelTablesNode.appendChild(nTable)
pdm.tableNodes[tableObject['code']] = nTable
nCode = pdm.ph.createElement('a:Code')
cCode = pdm.ph.createTextNode(tableObject['code'])
nCode.appendChild(cCode)
nTable.appendChild(nCode)
# tableObject['name'] = tableObject['code'] if tableObject.get('name') != None else tableObject['name']
# if tableObject.get('name') != None:
nName = pdm.ph.createElement('a:Name')
cName = pdm.ph.createTextNode(tableObject['code'] if tableObject.get('name') == None else tableObject['name'])
nName.appendChild(cName)
nTable.appendChild(nName)
if tableObject.get('comment') != None:
nComment = pdm.ph.createElement('a:Comment')
cComment = pdm.ph.createTextNode(tableObject['comment'])
nComment.appendChild(cComment)
nTable.appendChild(nComment)
nColumns = pdm.ph.createElement('c:Columns')
nTable.appendChild(nColumns)
@staticmethod
def updateTable(pdm, tableCode,tableObject=None):
"""
@breif 增加表
:param pdm: dom文档
:return:
"""
if tableObject == None:
print('待修改表不能为空且格式为{"name":,....}')
return False
tableNode = pdm.tableNodes[tableCode]
for key in tableObject.keys():
if key != 'code':
tf = False
for node in tableNode.childNodes:
if node.nodeName == 'a:' + key.capitalize():
node.removeChild(node.childNodes[0])
cName = pdm.ph.createTextNode(tableObject[key])
node.appendChild(cName)
tf = True
break
if tf == False:
tsNode = pdm.ph.createElement('a:' + key.capitalize())
tableNode.appendChild(tsNode)
cName = pdm.ph.createTextNode(tableObject[key])
tsNode.appendChild(cName)
@staticmethod
def delTable(pdm, tableCodeList=None):
"""
@breif 删除表
:param pdm: dom文档
:return:
"""
if tableCodeList == None or len(tableCodeList) == 0:
print('待删除表code列表不能为空')
return False
for tableCode in tableCodeList:
tableNode = pdm.tableNodes[tableCode]
tableNode.parentNode.removeChild(tableNode)
pdm.tableNodes.pop(tableCode)
keysTmp = []
for key in pdm.columnNodes.keys():
keysTmp.append(key)
for key in keysTmp:
if key.startswith(tableCode + '_'):
pdm.columnNodes.pop(key)
@staticmethod
def delColumn(pdm,tableCode, columnCodeList=None):
"""
@breif 删除列
:param pdm: dom文档
:return:
"""
if columnCodeList == None or len(columnCodeList) == 0:
print("待删除表code列表不能为空且格式如下['columnCode'...]")
return False
for columnCode in columnCodeList:
colunmNode = pdm.columnNodes[tableCode + '_' + columnCode]
colunmNode.parentNode.removeChild(colunmNode)
pdm.columnNodes.pop(tableCode + '_' + columnCode)
DDL解析操作脚本：
#!/usr/bin/python # -*- coding: utf-8 -*- # import re class DDLHandler(object): """ @preif DDL解析 """ ANNOTION = ['-','#'] LIST_DDL = ['CREATE TABLE','DROP TABLE','ALTER TABLE','COMMENT ON TABLE','COMMENT ON COLUMN'] list_ddl = [] list_ddl_line_join = [] fileName = 'ddl.sql' list_table_add = [] list_column_add = {} list_table_del = [] list_table_update = {} list_column_update = {} list_column_del = {} def __init__(self): self.init() self.joinLine() self.parseDdl() return def init(self): with open(self.fileName,'r',encoding='utf-8') as fl: for line in fl: if not re.match('^\s*('+'|'.join(self.ANNOTION)+')',line): line = re.sub('\s+',' ', line) self.list_ddl.append(line.upper().strip()) def joinLine(self): list_join = [] for line in self.list_ddl: if re.match('^\s*('+'|'.join(self.LIST_DDL)+')', line): if re.match('^\s*' + self.LIST_DDL[0], line): if re.match('.*\);\s*$', line): self.list_ddl_line_join.append(line) else: list_join.append(line) else: self.list_ddl_line_join.append(line) elif re.match('^\s*\);\s*$', line): list_join.append(line) self.list_ddl_line_join.append(' '.join(list_join)) list_join = [] else: list_join.append(line) def parseDdl(self): for sql in self.list_ddl_line_join: if re.match('^\s*'+self.LIST_DDL[0], sql): self.parseTableCreate(sql) elif re.match('^\s*'+self.LIST_DDL[1], sql): self.parseTableDel(sql) elif re.match('^\s*'+self.LIST_DDL[2], sql): self.parseColumnUpdate(sql) elif re.match('^\s*'+self.LIST_DDL[3], sql): self.parseTableComment(sql) elif re.match('^\s*'+self.LIST_DDL[4], sql): self.parseColumnComment(sql) def parseTableCreate(self,sql): regex = re.compile('^\s*CREATE\s+TABLE\s+(?P<tableCode>\w+)\s*\( (?P<columnCodes>.+)\);') match = regex.match(sql) tableCode = match.groupdict()['tableCode'] columnCodes = match.groupdict()['columnCodes'] columnCodes = columnCodes.split(',') tableObject = {'code':tableCode} self.list_table_add.append(tableObject) list = [] for column in columnCodes: regex = re.compile('^\s*(?P<columnCode>\w+)\s+(?P<dataType>\w+)(\((?P<length>\d+)\))?.*$') match = regex.match(column) length = match.groupdict()['length'] if match.groupdict().get('length') != None else '' dataType = match.groupdict()['dataType']+ ('('+match.groupdict()['length']+')' if match.groupdict().get('length') != None else '') data = {'code':match.groupdict()['columnCode'],'dataType':dataType,'length':length} list.append(data) self.list_column_add[tableCode] = list def parseTableDel(self, sql): regex = re.compile('^\s*DROP\s+TABLE\s+(?P<tableCode>\w+)\s*;') match = regex.match(sql) self.list_table_del.append(match.groupdict()['tableCode']) def parseColumnUpdate(self, sql): if re.match('^\s*ALTER\s+TABLE\s+\w+\s+ADD', sql): regex = re.compile('^\s*ALTER\s+TABLE\s+(?P<tableCode>\w+)\s+ADD\s*\((?P<columnCode>\w+)\s+(?P<dataType>\w+)(\((?P<length>\d+)\))?\);$') match = regex.match(sql) length = match.groupdict()['length'] if match.groupdict().get('length') != None else '' dataType = match.groupdict()['dataType'] + ('(' + match.groupdict()['length'] + ')' if match.groupdict().get('length') != None else '') data = {'code':match.groupdict()['columnCode'],'dataType':dataType,'length':length} if self.list_column_add.get(match.groupdict()['tableCode']) == None: self.list_column_add[match.groupdict()['tableCode']] = [data] else: self.list_column_add[match.groupdict()['tableCode']].append(data) elif re.match('^\s*ALTER\s+TABLE\s+\w+\s+DROP', sql): regex = re.compile('^\s*ALTER\s+TABLE\s+(?P<tableCode>\w+)\s+DROP\s+COLUMN\s+(?P<columnCode>\w+);$') match = regex.match(sql) if self.list_column_del.get(match.groupdict()['tableCode']) == None: self.list_column_del[match.groupdict()['tableCode']] = [match.groupdict()['columnCode']] else: self.list_column_del[match.groupdict()['tableCode']].append(match.groupdict()['columnCode']) elif re.match('^\s*ALTER\s+TABLE\s+\w+\s+MODIFY', sql): regex = re.compile('^\s*ALTER\s+TABLE\s+(?P<tableCode>\w+)\s+MODIFY\s*\((?P<columnCode>\w+)\s+(?P<dataType>\w+)(\((?P<length>\d+)\))?\);$') match = regex.match(sql) length = match.groupdict()['length'] if match.groupdict().get('length') != None else '' dataType = match.groupdict()['dataType'] + ('(' + match.groupdict()['length'] + ')' if match.groupdict().get('length') != None else '') data = {'code': match.groupdict()['columnCode'], 'dataType': dataType, 'length': length} if self.list_column_update.get(match.groupdict()['tableCode']) == None: self.list_column_update[match.groupdict()['tableCode']] = [data] else: self.list_column_update[match.groupdict()['tableCode']].append(data) def parseTableComment(self,sql): regex = re.compile("^\s*COMMENT\s+ON\s+TABLE\s+(?P<tableCode>\w+)\s+IS\s+'(?P<comment>.*)';$") match = regex.match(sql) if self.list_table_update.get(match.groupdict()['tableCode']) == None: self.list_table_update[match.groupdict()['tableCode']] = {'comment':match.groupdict()['comment']} else: self.list_table_update[match.groupdict()['tableCode']]['comment'] = match.groupdict()['comment'] def parseColumnComment(self,sql): regex = re.compile("^\s*COMMENT\s+ON\s+COLUMN\s+(?P<tableCode>\w+)\.(?P<columnCode>\w+)\s+IS\s+'(?P<comment>.*)';$") match = regex.match(sql) data = {'code': match.groupdict()['columnCode'], 'comment': match.groupdict()['comment']} if self.list_column_update.get(match.groupdict()['tableCode']) == None: self.list_column_update[match.groupdict()['tableCode']] = [data] else: self.list_column_update[match.groupdict()['tableCode']].append(data) if __name__ == '__main__': # ddl = DDLHandler() # text = 'CREATE TABLE STUDENTS( SNO CHAR(10), SNAME VARCHAR2(20), AGE NUMBER(3), SCHOOLDATE DATE, GRADE NUMBER(1) DEFAULT 1, CLAZZ NUMBER(2), EMAIL VARCHAR2(100) );' # ddl.parseTableCreate(text) # print(ddl.list_column_add) # text = "COMMENT ON COLUMN BANNER.NAME IS 'COMMENT';" # regex = re.compile("^\s*COMMENT\s+ON\s+COLUMN\s+(?P<tableCode>\w+)\.(?P<columnCode>\w+)\s+IS\s+'(?P<comment>.*)';$") # match = regex.match(text) # print(match.groupdict()) text = 'CREATE TABLE NOTE' print(re.sub('\s+',' ',text)) # if re.match('^\s*' + 'CREATE TABLE', text.replace('\s+',' ')): # print(True) # else: # print(False)DDL文件：--创建表结构 CREATE TABLE STUDENT( SNO CHAR(10), SNAME VARCHAR2(20), AGE NUMBER(3), SCHOOLDATE DATE, GRADE NUMBER(1) DEFAULT 1, CLAZZ NUMBER(2), EMAIL VARCHAR2(100) ); CREATE TABLE STUDENTS( SNO CHAR(10), SNAME VARCHAR2(20), AGE NUMBER(3), SCHOOLDATE DATE, GRADE NUMBER(1) DEFAULT 1, CLAZZ NUMBER(2), EMAIL VARCHAR2(100) ); --删除表结构 DROP TABLE STUDENT; DROP TABLE STUDENTS; --修改表的名称 RENAME BANK TO SXTBANK; --向表中添加注释 COMMENT ON TABLE STUDENT is 'banner'; -------------------修改表的结构 --添加列 ALTER TABLE STUDENT ADD(ADDRESS VARCHAR2(100)); ALTER TABLE STUDENT ADD(CLASS VARCHAR2(100)); --删除列 ALTER TABLE STUDENT DROP COLUMN ADDRESS; ALTER TABLE STUDENT DROP COLUMN AGE; ALTER TABLE STUDENTS DROP COLUMN SCHOOLDATE; --修改列 ALTER TABLE STUDENT MODIFY(EMAIL VARCHAR2(200)); ALTER TABLE STUDENT MODIFY(CLASS VARCHAR2(200)); --向列添加注释 COMMENT ON COLUMN STUDENT.sname is 'the name of employees';测试脚本：from PDMHandler import PDMHandler from DDLHandler import DDLHandler if __name__ == '__main__': ddl = DDLHandler() pmd = PDMHandler() #创建表 if len(ddl.list_table_add) != 0: for sql_table_add in ddl.list_table_add: PDMHandler.addTable(pmd, sql_table_add) #修改表 if len(ddl.list_table_update) != 0: for key in ddl.list_table_update.keys(): PDMHandler.updateTable(pmd, key, ddl.list_table_update[key]) #增加列 if len(ddl.list_column_add) != 0: for key in ddl.list_column_add.keys(): PDMHandler.addColumn(pmd, key, ddl.list_column_add[key]) #修改列 if len(ddl.list_column_update) != 0: for key in ddl.list_column_update.keys(): PDMHandler.updateColumn(pmd, key, ddl.list_column_update[key]) #删除列 if len(ddl.list_column_del) != 0: for key in ddl.list_column_del.keys(): PDMHandler.delColumn(pmd, key, ddl.list_column_del[key]) #删除表 if len(ddl.list_table_del) != 0: PDMHandler.delTable(pmd, ddl.list_table_del) #写入文件 with open('package_tmp.pdm', 'w',encoding = 'utf-8') as xmlfile: pmd.ph.writexml(xmlfile, addindent=' ' * 2, newl='\n', encoding='utf-8')测试通过，待改进
类似资料：