This function can be useful in any scenario, where the evaluation of a function call requires repeated evaluation of this same, or other functions. Often in such cases, the algorithm needs access to a general structure, containing the cached results of executing possibly many different functions applied on specific arguments-tuples. A unique key is needed under which to group all invocations of a specific function and then the mapping between their function-call arguments and the result of the call. The identity of a function fits exactly the requirements (uniqueness / one per function) for such a key.
This function enables applications to test whether two expressions or variables reference the same function. This may be useful, for example, to allow caching of function results to avoid repeated evaluation. The results of previous function invocations might be held in a map whose key is the function identity.
The function identity, by definition, is generated upon the creation of a function, is implementation-defined, and has meaning throughout the life of that function. It is not meaningful to store this value across different executions, because the identity given to a function in execution1 will generally be different from the identity, given to it in execution2.The function identity, by definition, is generated upon the creation of a function. It is not meaningful to store or compare the result across different execution scopes, because the string used to represent the function identity will generally vary from one execution scope to another.The function identity, by definition, is generated upon the creation of a function, is implementation-defined, and has meaning throughout the life of that function. It is not meaningful to store this valueor compare the result across different executionsexecution scopes, because the identity given to a function instring used to represent the function identity execution1 will generally be different from the identity, given to it in vary from one execution scope to anotherexecution2.
The function identity, being a string, can be used as a key in a map, thus making it possible to map a particular function to a sequence of items.
The result of an expression such as function-identity(abs#1) eq function-identity(abs(?)) may be either true or false, because it is implementation-dependent whether abs#1 and abs(?) return the same function.