I feel like this might be a crazy question, and if anyone has a better idea of how to do it, by all means please let me know.
I have a .Net C# project at work that validates certain tables against the company standards they must adhere to. The tables can be on SQL Server or DB2. When a table is initially added to the program, I gather the metadata/information about the table using select statements off of either information_schema.columns for sql, or syscat.columns for db2. Once I get that data, I store it in a table (let’s call it all_table_information) on sql server containing that same information for every table that is validated in the program.
If the table is a SQL table, I can run this query (of course restricting it to only the columns from the table I want):
insert into [all_table_information] (table_id, column_name, data_type, max_char_length) select table_id, column_name, data_type, character_maximum_length from information_schema.columns where ...restrict to needed table... And then execute that in a SqlCommand. But if it is a DB2 table, I have to run this query (restricted again to the columns I need):
select tabschema, tabname, colname, typename, length from syscat.columns where ...restrict to needed table... Then get a DataReader with the results and loop through it inserting each row using:
while (dr.Read()) { insert into [all_table_information] (table_id, column_name, data_type, max_char_length) values (...''' + dr['whatever'] + '', '' + ....) ...execute sql here... } dr.Close(); Doing it this way works, but it is slow, so I was just wondering if there was anyway to do this in one statement? I know you could save the db2 table information to a data table as well, but can you run a query against a data table directly?
Thanks, Ryan
You probably want to use SqlBulkCopy: MSDN
It is much faster than doing individual inserts for each row.