Skip to main content

Output data structure

We want to load into the database, California clients living in a couple of Counties only: Orange and Los Angeles counties.

The table structure is slightly different, therefore the data expected to be loaded into the DB table should have the following structure:

  • Key (key, Type: Integer)

  • Name (Type: String, max. length: 40)

  • Address (Type: String, max.length: 40)

  • County (Type: String, max. length:40)

In order to load this table, we will need to use the following mapping process:

The Key column is fed with an auto-incremented integer.

The Name column is filled out with a concatenation of first and last names.

The Address column data comes from the equivalent Address column of the input file, but supports a upper-case transformation before the loading.

The County column is fed with the name of the County where the city is located using a reference file which will help filtering Orange and Los Angeles counties' cities.

Did this page help you?

If you find any issues with this page or its content – a typo, a missing step, or a technical error – please let us know!