What are some common sources of side effects in functional programming?

1 Input and output

One of the most obvious sources of side effects is input and output (I/O), such as reading from a file, writing to a database, or interacting with the user. I/O functions are inherently impure, because they depend on the state of the external world, which can change at any time. Moreover, they can also modify the state of the world, which can affect other functions. To deal with I/O in a functional way, we can use techniques such as monads, which encapsulate the impure operations and provide a way to compose them in a pure manner.

2 Mutable data structures

Another common source of side effects is mutable data structures, such as arrays, objects, or variables that can be modified in place. Mutable data structures can introduce side effects when they are shared by multiple functions, which can alter their contents and affect each other's behavior. For example, if we have a function that sorts an array in place, and another function that uses the same array, we can end up with unexpected results. To avoid this, we can use immutable data structures, which are data structures that cannot be changed once created. Immutable data structures can be copied, modified, and returned as new values, without affecting the original ones.

4 Randomness and time

Another common source of side effects is randomness and time, which are factors that can introduce variability and uncertainty in the program. Randomness and time functions are impure, because they depend on the state of the system or the environment, which can change at any time. Moreover, they can also affect the state of the system or the environment, which can affect other functions. For example, if we have a function that generates a random number, and another function that uses that number, we can have different outcomes depending on when and how often the functions are called. To deal with randomness and time in a functional way, we can use techniques such as seeds, which provide a way to control and reproduce the random values, or pure functions, which abstract the time-dependent logic and return a value based on a given input.

5 Exceptions and errors

Another common source of side effects is exceptions and errors, which are situations that disrupt the normal flow of the program and require special handling. Exceptions and errors functions are impure, because they depend on the state of the program or the environment, which can cause failures or crashes. Moreover, they can also modify the state of the program or the environment, which can affect other functions. For example, if we have a function that divides two numbers, and another function that uses the result, we can have a problem if the divisor is zero, which can throw an exception or return an error. To deal with exceptions and errors in a functional way, we can use techniques such as option types, which represent the possibility of a value or an error, or error handling functions, which provide a way to handle and recover from errors in a pure manner.