我正在使用外部实用程序来备份我的MS-SQL dbs，它会备份我的dbs并在恢复它时会给我db文件（.mdf，.ldf）文件。
现在，我需要的是，验证Restored db文件（.mdf＆amp; .ldf文件）是否与原始db文件相同或不使用$(document).ready(function(e) {
$('#loginButton').click(function(e) {
$.post('../php/login.php', {
functions: 'userLogin',
username: $('#username').val(),
userpass: $('#userpass').val()
}).done(function( data ) {
alert("Data Loaded: " + data);
});
});
});
进行自动化。是否有任何实用程序可以与python集成db比较？
我正在调查pymssql模块，但不确定，我是否可以使用此模块比较恢复的数据库...
答案 0 :(得分：0)
我查看了pymssql。 pymssql的作用是，只执行传递给它的SQL查询并返回相同的输出。但是，安装pymssql（在Windows上）确实是一项非常困难的任务。因此，我更喜欢使用SQL Server的内置pymssql实用程序来执行我的SQL查询，而不是使用SQLCMD。
因此，我使用os.system()的{{1}}来通过python执行我的SQL查询。因此，我通过python将我的SQL查询写入.sql文件，然后将其传递给SQLCMD并将输出重定向到另一个文本文件。
首先，我使用SQLCMD收集了给定数据库中的表列表。然后，我逐个查询所有表，并将它们的数据放入一个单独的文本文件SQLCMD中。对于还原的db，我也遵循相同的过程并将数据转储到另一个文本文件(Original_DB_Data.txt)。代码段如下： -
(Restored_DB_Data.txt)如果您希望改进 import os
def get_table_list_in_DB(Instance_name, DB_name):
query = 'SELECT NAME from [%s].sys.tables' %(DB_name)
input_file = "C:\\SQL_Data\\SQLQuery_table_list.sql"
output_file = "C:\\SQL_Data\\Table_list_DB-%s_Output.txt" %(DB_name)
with open(input_file, 'w') as f:
f.write(query)
command = 'sqlcmd -S %s -i %s -o %s' %(Instance_name, input_file, output_file)
os.system(command)
os.remove(input_file) # to delete the created .sql file
return output_file
def get_db_data(Instance_name, DB_name):
output_file = "C:\\SQL_Data\\DB_Data-%s_Output.txt" %(DB_name)
table_list = get_table_list_in_DB(Instance_name, DB_name)
flag = 0
with open(table_list, 'r') as f1, open(output_file, 'a') as f2:
for lines in f1:
if re.match("^\s", lines):
flag = 0
if flag:
table_data = get_table_data(Instance_name, DB_name, table_name=lines.strip())
with open(table_data, 'r') as f3:
f2.write("##################################" + '\n')
f2.write('\t' +lines.strip() + '\n')
f2.write("##################################" + '\n')
f2.write(f3.read())
os.remove(table_data)
if re.match("^----+", lines):
flag = 1
return output_file
def get_table_data(Instance_name, DB_name, table_name):
input_file = "C:\\SQL_Data\\SQLQuery_table_data.sql"
output_file = "C:\\SQL_Data\\Table_data_%s_Output.txt" %(table_name)
query = "SELECT * from [%s].dbo.[%s]" %(DB_name, table_name)
with open(input_file, 'w') as f:
f.write(query)
command = "sqlcmd -S %s -i %s -o %s" %(Instance_name, input_file, output_file)
os.system(command)
os.remove(input_file)
return output_file
def compare_DB_Data(DB_Detail1=[], DB_Detail2=[]):
get_db_data(Instance_name=DB_Detail1[0], DB_name=DB_Detail1[1])
get_db_data(Instance_name=DB_Detail2[0], DB_name=DB_Detail2[1])
data_DB1 = "C:\\SQL_Data\\DB_Data-%s_Output.txt" %(DB_Detail1[1])
data_DB2 = "C:\\SQL_Data\\DB_Data-%s_Output.txt" %(DB_Detail2[1])
with open(data_DB1, 'r') as f1, open(data_DB2, 'r') as f2:
if f1.read() == f2.read():
print "Data of both DB Matches"
else:
print "Data of both DB Varies"
方法以获得确切的差异，那么您可以按如下方式重写它，这会将差异转储到另一个文本文件中，以后可以请参阅，以便检查究竟有什么区别： -
compare_DB_Data()答案 1 :(得分：0)
我故意添加这个答案。只有在SQL服务器所在的同一台机器上运行时，上述答案才有效。而且，早些时候我们使用的是文本文件，它不能确保两个dbs之间的完全比较。因此，这个新的方法是基于字典的方法来避免上述问题。
即使您的SQL服务器和代码位于不同的计算机上，以下代码仍然有效。但是，为此，您需要在存在SQL Server的计算机上启用WINRM模块。 WINRM是一个窗口的内置模块，用于在Windows机器之间进行通信。要在计算机上启用WINRM，您需要在SQL计算机的命令行中运行以下命令。
winrm qc -q
winrm set winrm/config/client/auth @{Basic="true"}
winrm set winrm/config/service/auth @{Basic="true"}
winrm set winrm/config/service @{AllowUnencrypted="true"}
这将在您的计算机上启用WINRM模块。现在我们需要pywinrm python模块来与我们的远程SQL主机进行通信。有关安装pywinrm的信息，请参阅this链接。
所以，首先我们将收集数据库中存在的表列表（使用方法get_table_list_in_DB。将这些表存储到文本文件中。此列表将用于收集表模式和表数据库中存在的所有表的数据。逐个从文件中读取每个表名，并查询表模式和表数据。
get_db_Schema()方法读取上面返回的表列表并在内部调用get_table_Schema()方法，收集数据库中所有表的架构详细信息。为每个表返回的模式存储为字典键值对，其中表名是键，返回的模式是值。因此，整个数据库模式是一个字典，每个表为key，其模式为value。
以同样的方式，get_db_data()和get_table_data()有效。 get_db_data()收集每个表的数据并存储在字典中，表名为key，返回的数据为value。
因此，为了比较两个数据库，我们将数据库详细信息（数据库名称和实例名称）传递给compare_DB_Data()方法，后者又调用两个数据库的get_db_Schema()并进行比较。如果两个数据库的模式匹配，则in将为两个数据库调用get_db_data()方法并匹配它们。因为，我们以键值对的形式存储了表模式和表数据，所以如果DB1中的每个键（对于模式和数据）在两个DB中都具有相同的值，那么我们可以确保两者都有我们的DB是一样的。
如果发现两个DB的架构有任何差异，则会将这些差异添加到schema1_diff_schema2 schema2_diff_schema1字典中。同样，数据差异也会添加到data1_diff_data2和data2_diff_data1字典中。
以下是相同的代码： -
class SQL_Compare_DB(object):
def __init__(self, SQL_Host_IP, auth):
# Append to path, in case not present
sys.path.append(r"C:\Program Files\Microsoft SQL Server\110\Tools\Binn")
# Create session with SQL host
self.session = winrm.Session(SQL_Host_IP, auth)
# We need a directory where the db files will be stored
if not os.path.exists("C:\SQL_Data"):
os.mkdir("C:\SQL_Data")
else:
os.system("RMDIR /S /Q C:\SQL_Data")
os.system("MKDIR C:\SQL_Data")
def get_table_list_in_DB(self, Instance_name, DB_name):
'''
Returns the list of table in the given database
'''
output_file = "C:\\SQL_Data\\Table_list_DB-%s_Output.txt" %(DB_name)
query = 'SELECT NAME from [%s].sys.tables' %(DB_name)
command = 'sqlcmd -S "%s" -Q "%s"' %(Instance_name, query)
execute_query = self.session.run_cmd(command)
if execute_query.std_err:
print "Error in command execution :- ", execute_query.std_err
return False
with open(output_file, 'w') as f:
f.write(execute_query.std_out)
return output_file
def get_db_Schema(self, Instance_name, DB_name):
'''
Get the schema of all tables in the Database
'''
table_list = self.get_table_list_in_DB(Instance_name, DB_name)
flag = 0
db_schema = {}
with open(table_list, 'r') as f1:
for lines in f1:
if re.match("^\s", lines):
flag = 0
if flag:
table_schema = self.get_table_Schema(
Instance_name, DB_name, table_name=lines.strip())
db_schema.update(table_schema)
if re.match("^----+?-", lines):
flag = 1
print db_schema
return db_schema
def get_table_Schema(self, Instance_name, DB_name, table_name):
'''
Get the table schema
'''
table_schema = {}
query = """
SELECT ORDINAL_POSITION, COLUMN_NAME, DATA_TYPE,
CHARACTER_MAXIMUM_LENGTH, IS_NULLABLE
FROM [%s].INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME = '%s'""".replace('\n','') %(DB_name, table_name)
command = 'sqlcmd -S "%s" -Q "%s"' %(Instance_name, query)
execute_query = self.session.run_cmd(command)
table_schema[table_name] = execute_query.std_out
if execute_query.std_err:
print "Error in command execution :- ", execute_query.std_err
return False
return table_schema
def get_db_data(self, Instance_name, DB_name):
'''
From the given DB, it fetchs the data of all existing
tables and append those in a single text file
'''
db_data = {}
table_list = self.get_table_list_in_DB(Instance_name, DB_name)
flag = 0
with open(table_list, 'r') as f1:
for lines in f1:
if re.match("^\s", lines):
flag = 0
if flag:
table_data = self.get_table_data(
Instance_name, DB_name, table_name=lines.strip())
db_data.update(table_data)
if re.match("^----+?-", lines): # after the line ------ the tabel data are printed
flag = 1
print db_data
return db_data
def get_table_data(self, Instance_name, DB_name, table_name):
'''
Get the data for the given table
'''
table_data = {}
query = "SELECT * from [%s].dbo.[%s]" %(DB_name, table_name)
command = 'sqlcmd -S "%s" -Q "%s"' %(Instance_name, query)
execute_query = self.session.run_cmd(command)
table_data[table_name] = execute_query.std_out
if execute_query.std_err:
print "Error in command execution :- ", execute_query.std_err
return False
return table_data
def compare_DB_Data(self, DB_Detail1=[], DB_Detail2=[]):
'''
Take detail of two DBs as input and gets the detailed DB data
The data collected are dumped into one dictionary and at last
both the dictionary are compared.
Arguments :-
DB_Detail1 :- a list of instance name and DB name for original DB
DB_Detail2 :- a list of instance name and DB name for Restored DB
e.g :- DB_Detail1 = ['Instance_name', 'Database_name']
'''
# Compare schema
db_schema1 = self.get_db_Schema(
Instance_name=DB_Detail1[0], DB_name=DB_Detail1[1])
db_schema2 = self.get_db_Schema(
Instance_name=DB_Detail2[0], DB_name=DB_Detail2[1])
schema1_diff_schema2 = {}
schema2_diff_schema1 = {}
print db_schema1
print db_schema2
set_current, set_past = set(db_schema1.keys()), set(db_schema2.keys())
intersect = set_current.intersection(set_past)
added = set_current - intersect
removed = set_past - intersect
changed = set(k for k in intersect if db_schema2[k] != db_schema1[k])
unchanged = set(k for k in intersect if db_schema2[k] == db_schema1[k])
print added,removed,changed,unchanged
[schema1_diff_schema2.update(i) for i in [{m :db_schema1[m]} for m in added ]]
[schema1_diff_schema2.update(i) for i in [{m :db_schema1[m]} for m in changed]]
[schema2_diff_schema1.update(i) for i in [{m :db_schema2[m]} for m in removed]]
[schema2_diff_schema1.update(i) for i in [{m :db_schema2[m]} for m in changed]]
if added == set([]) and removed == set([]) and changed == set([]):
print "Schema of both DB Matches"
else:
print "Schema of both DB Varies"
# Compare data
data_DB1 = self.get_db_data(
Instance_name=DB_Detail1[0], DB_name=DB_Detail1[1])
data_DB2 = self.get_db_data(
Instance_name=DB_Detail2[0], DB_name=DB_Detail2[1])
data1_diff_data2 = {}
data2_diff_data1 = {}
set_current_data, set_past_data = set(data_DB1.keys()), set(data_DB2.keys())
intersect = set_current_data.intersection(set_past_data)
added = set_current_data - intersect
removed = set_past_data - intersect
changed = set(k for k in intersect if data_DB2[k] != data_DB1[k])
unchanged = set(k for k in intersect if data_DB2[k] == data_DB1[k])
print added,removed,changed,unchanged
[data1_diff_data2.update(i) for i in [{m :data_DB1[m]} for m in added ]]
[data1_diff_data2.update(i) for i in [{m :data_DB1[m]} for m in changed]]
[data2_diff_data1.update(i) for i in [{m :data_DB2[m]} for m in removed]]
[data2_diff_data1.update(i) for i in [{m :data_DB2[m]} for m in changed]]
print "Diff DB1 vs DB2 :- ", data1_diff_data2
print "Diff DB1 vs DB2 :- ", data2_diff_data1
if added == set([]) and removed == set([]) and changed == set([]):
print "Data of both DB Matches"
else:
print "Data of both DB Varies"