I sometimes write functions that make the assumption that some arguments cannot occur. If they do, this is a bug and I fail:
let foo = function
| 0 -> ()
| _ -> failwith "foo: bad argument"
If I rename the function later on, I have to remember to also change the string. Is there a way to do this in a more systematic manner? My mind wanders around solutions like
| _ -> failwith (FUNCTION_NAME ^ ": bad argument")
where FUNCTION_NAME is a string variable that the compiler or interpreter will instantiate. But I have no idea whether something like this even works in OCaml. If not, is there a best practice?
There is a set of values available for debugging and error reporting.
OCaml 4.12 introduced __FUNCTION__:
val __FUNCTION__ : string
__FUNCTION__returns the name of the current function or method, including any enclosing modules or classes.
They might be useful if you don't want to use assert false as suggested by @SteveVinoski.
__LOC__ : string
__FILE__ : string
__LINE__ : int
__MODULE__ : string
__POS__ : string * int * int * int
There are also fancier forms that you can use to wrap an expression to determine its extent in the source.
These are documented in the Stdlib module.
I think you should try as much as possible to use more specific types so that the compiler can prevent you from calling your functions with invalid input.
If you really don't find a way to tell the type system what you want, try harder, and if you really can't, then use assert which as Steve told you will give you some precious debugging information (file and line number are much more useful than function name since chances are high your function doesn't have a name).
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With