Here is where the problems start as the headers will force to read all the data in String format. Then depending on the complexity of the data this is something you will have to deal with afterwards. Your data will look something like this:
Then you can fix it easily with the following workflow…
Thank you for your reply! However I in a number of excel files I have the headers start in col C row 5 but have some text in col b row 2.
If I try your solution, the splitter will use the text in col b row 2 as header instead of the actual headers that are in col c row 5. Would you know how it implement a work-around for this issue? The “random” text in col b row 2 is usually at most one cell.
I have modified the workflow with his method of detecting were the data starts, fitting it into a loop, that search for a field name that can be fitted at any column order… to avoid the creativeness of our excel reporter friends.