Working with Missing Values

There are times when data may not have a value. For example, in an application that is designed to capture a user's address the variable holding the address will be empty until the user has filled in their address. Another example may be a user that doesn't have a middle name. Variables must always have a value, so The Null Data Type is used to indicate that a certain piece of data is missing.

If you were to store the address a user enters in a variable called address and access that variable before the user has entered their address you would end up with the value NULL.

address => NULL

Taking this example a bit further, if you were to access a property from the address such as zip_code you would also end up with the value NULL.

address.zip_code => NULL

If some of your data may be missing, it is a good idea to check if it is NULL before using it. For example, if you wanted to display the zip_code field to the user and there is no value it would be better to display a helpful message instead.

IF(address.zip_code = NULL, "Zip code unknown.", address.zip_code)

Working with Empty Values

In the previous example, what would you display if the zip_code is the empty string ("")?. Well, we could of course also check for that.

  address.zip_code = NULL OR = "",  
  "Zip code unknown.",  

However, it is much more convenient to use the ISEMPTY function, which returns FALSE for both "", as well as NULL.

IF(ISEMPTY(address.zip_code), "Zip code unknown.", address.zip_code)

There is a companion ISNOTEMPTY function, so we could accomplish the same by changing the ordering.

IF(ISNOTEMPTY(address.zip_code), address.zip_code, "Zip code unknown.")

This situation is not unique to strings, you may want to handle an empty list the same way you treat NULL. The following values are considered to be empty.

  • The null value: NULL
  • An empty list: []
  • An empty object: {}
  • An empty string: ""

Working with Values that are Considered Falsy

The IF function, the WHERE clause of the Query Expression, and the filtering Path Bindings all examine their values in terms of whether they are falsy or not. For example, the empty string ("") is falsy, so we did not need to use they ISNOTEMPTY function at all in our previous example.

IF(address.zip_code, address.zip_code, "Zip code unknown")

The following values are considered falsey

Working with Values of an Unknown Type

You may not always know that a value will be of a certain type. For example, even though you asked the user to enter a phone number, they may have entered something else like "Hello Airkit!". Not only, can you not call a phone number like "Hello Airkit", a phone number must be in a very specific format in order to be called (or sent an SMS). The ISPHONE function will return TRUE when a string represents a phone number that can be called. 

Several types have a corresponding function that tells you whether or not a value is of that type.