Column reference example:

Output: Returns the number of characters in the value in column MyName.

String literal example:

Output: Returns the value 12.

column_string

Name of the column or string constant to be searched.

• Missing string or column values generate missing string results.
• String constants must be quoted ('Hello, World').
• Multiple columns and wildcards are not supported.

Example - Fixed Length Strings

Source:

Your product identifiers follow a specific structure that you'd like to validate in your recipe. In the following example data, the productId column should contain values of length 6.

You can see that there is already a column containing validation errors for the ProductName column. Values in the ProductId column that are not this length should be flagged in a new column. Then, you should merge the two columns together to create a ValidationError column.

Transformation:

To validate the length of the values in ProductId, enter the following transform. Note that the as parameter enables you to rename the column as part of the transform.

The dataset now looks like the following:

You can blend the two error columns into a single DataValidationErrors error column using the following merge transform. Note again the use of the as parameter:

To clean up the data, you might want to do the following, which trims out the whitespace in the DataValidationErrors column and removes the two individual error columns:

Results:

The final dataset should look like the following: