Backreferences

Backreferences are particularly effective when you need to detect a specific pattern within text data and reuse part of it to convert it into a new format.

For example, let’s consider converting the text “Hello World” to “World Hello”.

By using the regular expression ([^ ]+) ([^ ]+), you can capture the strings before and after the space as individual groups. Then, by using the backreference \2 \1, you can rearrange the captured groups in reverse order to get “World Hello”.

To explain ([^ ]+) in more detail, it consists of the following components, forming a group that matches a sequence of any characters except spaces:

• ( ) - Creates a group
• [^ ] - Matches any single character except a space
• + - Matches one or more repetitions of the preceding pattern

Now, let’s try this with customer data.

For instance, suppose you have customer name data like the following, and you want to convert the format from “Last Name, First Name” to “First Name, Last Name”.

From the column header menu of customer_name, select “Work with Text Data,” then “Replace,” and then “Text.”

The “Work with Text Data” dialog will appear; check the “Regular Expression” box.

Specify ([^ ]+) ([^ ]+) as the source (pattern to find) and try specifying \2 as the replacement. This allows you to extract “First Name,” which is the second group in the “Last Name, First Name” sequence.

By specifying \2 \1 as the replacement, you can see that the order is successfully swapped to “First Name, Last Name.”

By clicking “Run,” you have successfully converted the data from the “Last Name, First Name” format to the “First Name, Last Name” format.

Video