Question

Give at least two examples of situations in which hashing is not appropriate.

Short answer

Hashing is not appropriate when the data needs to be retrieved in the order it was originally input (e.g., for time-series data), or when the original data needs to be recovered from the hashed output (e.g., for encryption).

Step-by-step solution

Situation 1 - Original Order Retrieval

If the order in which data was originally entered or recorded matters for the application at hand, using a hash function could disrupt that order. As hash functions map input data to numerical values in a deterministic but seemingly random way, there's no way to reverse-engineer the input's original order from the hashed output. For instance, consider an application which records the time-series data (e.g. stock prices). If this data is hashed, the time-series information will be lost.

Situation 2 - Reversible Data

If you need to retrieve the original data from the hashed output, hashing is not appropriate. Although hash functions are deterministic, they are designed to be difficult to reverse. In other words, if you know the output of a hash function, it should still be computationally infeasible to determine what the input was. This is a feature, not a bug, for many applications (for example, storing password data). But for applications such as encryption where you want to be able to recover the original data by reversing the action, hashing is inappropriate. Encryption requires using a deterministic function that's easily reversible with the right key.