SQLBrowseConnect, like SQLDriverConnect, uses a connection string. However, by using SQLBrowseConnect, an application can construct a complete connection string at run time. This allows the application to do two things:
The application calls SQLBrowseConnect and passes a connection string, known as the browse request connection string, that specifies a driver or data source. The driver returns a connection string, known as the browse result connection string, that contains keywords, possible values (if the keyword accepts a discrete set of values), and user-friendly names. The application builds a dialog box with the user-friendly names and prompts the user for values. It then builds a new browse request connection string from these values and returns this to the driver with another call to SQLBrowseConnect.
Because connection strings are passed back and forth, the driver can provide several levels of browsing by returning a new connection string when the application returns the old one. For example, the first time an application calls SQLBrowseConnect, the driver might return keywords to prompt the user for a server name. When the application returns the server name, the driver might return keywords to prompt the user for a database. The browsing process would be complete after the application returned the database name.
Each time SQLBrowseConnect returns a new browse result connection string, it returns SQL_NEED_DATA as its return code. This tells the application that the connection process is not complete. Until SQLBrowseConnect returns SQL_SUCCESS, the connection is in a Need Data state and cannot be used for other purposes, such as to set a connection attribute. The application can terminate the connection browsing process by calling SQLDisconnect.