I dint know many OO languages that don’t have a useless toString on string types. It mostly seems to be a result of using a generic string-able type that’s implemented to add toString() in a standardised way.
Calling toString on a string is practically a no-op anyway.
I dint know many OO languages that don’t have a useless toString on string types
Okay, fair enough. Guess I never found about it because I never had to do it… JS also allows for “test string”.toString() directly, not sure how it goes in other languages.
Java would be “test string”.toString(). C# has “test string”.ToString(). Python has str(“test string”) (as str() is Python’s toString equivalent). Rust has String::from(“test string”).to_string().
That’s just from the top of my head. I’m sure there’s more.
Edit: actually, I think Rust’s to_string() may not be entirely useless, I think it may be used as a consuming placeholder for clone()? Not sure how that would be useful, but it’s not a complete no-op at least.
It’s also incredibly useful as a failsafe in a helper method where you need the argument to be a string but someone might pass in something that is sort of a string. Lets you be a little more flexible in how your method gets called
I dint know many OO languages that don’t have a useless toString on string types.
Well, that’s just going to be one of those “it is what it is” things in an OO language if your base class has a toString()-equivalent. Sure, it’s probably useless for a string, but if everything’s an object and inherits from some top-level Object class with a toString() method, then you’re going to get a toString() method in strings too. You’re going to get a toString() in everything; in JS even functions have a toString() (the output of which depends on the implementation):
In a dynamically typed language, if you know that everything can be turned into a string with toString() (or the like), then you can just call that method on any value you have and not have to worry about whether it’ll hurl at runtime because eg. Strings don’t have a toString because it’d technically be useless.
Same is true for JavaScript’s namesake, Java; Object has a toString method, so everything but primitives (int, long, etc.) must have a toString method (and primitives sort of have one too in a roundabout way).
I think JavaScript’s toString also serves another function, namely to have some form of fallback when doing operations on what should be incompatible types. [] + “”, for instance; JavaScript will call toString() to do type conversion when the nearest matching type is a String.
Everything that’s an Object is going to either inherit Object.prototype.toString() (mdn) or provide its own implementation. Like I said in another comment, even functions have a toString() because they’re also objects.
A String is an Object, so it’s going to have a toString() method. It doesn’t inherit Object’s implementation, but provides one that’s sort of a no-op / identity function but not quite.
So, the thing is that when you say const someString = “test string”, you’re not actually creating a new String object instance and assigning it to someString, you’re creating a string (lowercase s!) primitive and assigning it to someString:
In Javascript, primitives don’t actually have any properties or methods, so when you call someString.toString() (or call any other method or access any property on someString), what happens is that someString is coerced into a String instance, and then toString() is called on that. Essentially it’s like going new String(someString).toString().
Now, what String.prototype.toString() (mdn) does is it returns the underlying string primitive and not the String instance itself:
Why? Fuckin beats me, I honestly can’t remember what the point of returning the primitive instead of the String instance is because I haven’t been elbow-deep in Javascript in years, but regardless this is what String’s toString() does. Probably has something to do with coercion logic.