An efficient way to read from a text file into a database table

Question

I have a log file. The file is structured. Each line is delimeted by commas so that it looks like a table. I have to read from this file and transport the content to a database table. I want to treat the first 9 commas as delimeters and the rest ones as just a content of the tenth column. So, there can be more than 9 commas in a line but the tenth and the next ones shoudn't be taken as a delimeter. I actually know how to do it by iterating through the characters and checking occurance of commas. But I don't want it be inefficient. maybe there's a better way? What would be the most accurate way to accomplish this? Either C# or Delphi is OK. For Oracle I'll probably use Oracle but SQL Server is also an option.