KexiDataMigrationAndSharing / GeneralHints

1. Establish connection using Connection object and use its methods, like:

 Cursor* executeQuery( const QString& statement, uint cursor_options = 0)
 Cursor* prepareQuery( const QString& statement, uint cursor_options = 0)
 bool querySingleRecord(const QString& sql, KexiDB::RowData &data)

2. Why is there redundancy in the Migration framework compared to KexiDB, e.g. why there is KexiMigrate::drv_tablenames() if we have KexiDB::Connection::tableNames()?

Note that there is one important difference: For example KexiDB::Connection::tableNames() uses KexiDB::Connection::objectNames() which in turn, uses SQL. It is not a rule that the source of the migration is SQL-compatible. It can be a structured text file, e.g. XML if we implemente that. We already have MS Access migration driver, and it does not use SQL for querying about structure or data - mdbtools call are used instead. Thus we were able to reimplement the tableNames method completely.