I've written a generic database class which can be called to carry out common database (CRUD) operations to save re-writing the ADO.NET code in multiple solutions. To make this flexible, there are a number of constructor overloads based on the different database authentication types and instance types etc. The class is as follows:
class Database
{
// default instance with Windows authentication
// constructor 1
public Database(string server, string database, bool persistSecurityInfo)
{
_server = server;
_database = database;
_persistSecurityInfo = persistSecurityInfo;
_integratedSecurity = "True";
_connectionString = "Data Source=" + server + ";Initial Catalog=" + database + ";Persist Security Info=" + persistSecurityInfo.ToString() + ";Integrated Security=True";
}
// named instance using Windows authentication
// constructor 2
public Database(string server, string instance, string database, bool persistSecurityInfo) : this(server, database, persistSecurityInfo)
{
_instance = instance;
_integratedSecurity = "True";
_connectionString = "Data Source=" + server + "\\" + instance + ";Initial Catalog=" + database + ";Persist Security Info=" + persistSecurityInfo.ToString() + ";Integrated Security=True";
}
// default instance with SQL authentication
// constructor 3
public Database(string server, string database, bool persistSecurityInfo, string userName, string password) : this(server, database, persistSecurityInfo)
{
_userName = userName;
_password = password;
_integratedSecurity = "False";
_connectionString = "Data Source=" + server + ";Initial Catalog=" + database + ";Persist Security Info=" + persistSecurityInfo.ToString() + ";User ID=" + userName + ";Password=" + password;
}
// named instance with SQL authentication
// constructor 4
public Database(string server, string instance, string database, bool persistSecurityInfo, string userName, string password) : this(server, database, persistSecurityInfo, userName, password)
{
_instance = instance;
_integratedSecurity = "False";
_connectionString = "Data Source=" + server + "\\" + instance + ";Initial Catalog=" + database + ";Persist Security Info=" + persistSecurityInfo.ToString() + ";User ID=" + userName + ";Password=" + password;
}
private string _server;
private string _instance;
private string _database;
private bool _persistSecurityInfo;
private string _userName;
private string _password;
private string _integratedSecurity;
private string _connectionString;
private string _query;
//CRUD Methods here
}
I have written a console application which is writing to a database. When the application is executed, the user provides some command line switches.
Some of the switches are as follows (There are others relating to the program's operation which I have not included here):
- /s : database server name
- /i : database instance name
- /d : database name
- /n : integrated security (True or False)
- /u : db Username
- /p : db Password
/i, /u and /p are optional (EG if an instance name isn't supplied, the program assumes it is to connect to a default instance on /s)
Therefore, I need the program to, at run time decide which constructor to call based on which arguments have been provided.
pseudo example here
Class Program
{
static void Main(string[] args)
{
foreach (string arg in args[])
{
//code to work out which parameters have been provided here and adds them to array. Also other code which checks integrity such as ensuring there is no username without a password and vice versa etc.
string[] suppliedParameters;
//if there is a /i , /u , /p parameters, use constructor 4
//if there is a /u and /p but no /i, use constructor 3
//if there is an /i but no /u or /n use constructor 2
//if there is no /i, /u or /n, use constructor 1
}
}
}
I know I can use reflection to execute the relevant constructor and that I could achieve selection of the constructor using a switch statement in the Main method which carries out the tests in the logic above but am just wondering if there is a maybe a more elegant way to do this?