Transform csv data from rows to columns and rows data frame

I am writing a program in Python to take csv data, put it in a data frame as is and then change that dataframe so that there are two MySQL tables created. I can do most of it except structure the data change. The data is as follws:

enter image description here

What I need to do is put unqiue Symobol values, description, currency and UOM into a one table and then link by an ID to a timeseries table that will have Date, Low, High, Close and Index. The other columns can be ignored and the first header row ignored.

The results will be that I will insert into MySQL tables this part I am ok with I just can work out how to link the distinct or unqiue codes back to the timeseries data.