What's the trick here?
['1', '2', '3'].map(parseInt); // While one could expect [1, 2, 3] // The actual result is [1, NaN, NaN]
parseInt is often used with one argument, but takes two. The first is an expression and the second is the radix. To the callback function, Array.prototype.map passes 3 arguments: the element, the index, the array. The third argument is ignored by parseInt, but not the second one, hence the possible confusion.
Neither of these are doing anything wrong on their own[1]. When you go to map something you want to be able to access the index sometimes. When you parse an int, you sometimes want to be able to specify a base. Maybe in a strongly typed language you might use an enum with valid bases instead of an int, but that doesn't make sense in a dynamically typed language[2]. Maybe you use different function names if you want to parse an int in decimal vs an arbitrary base, I can't argue with that, but overloaded functions/methods are the norm and not the exception in most languages I've used.
[1]: Except parseInt's default radix which is ambiguous and the cause of countless bugs. But this bug is because we ARE specifying it.
[2]: And no, being dynamically typed is not inherently bad design.
Actually, to be honest will NaN even launch an exception? Does 1 + NaN cause an exception or does it propagate more NaNs all over the code?
Imagine you are debugging a 10000 line function that's suppose to return an int, but instead returns a NaN. Where did this error occur in the 10000 lines? Sure, I can say all kinds of stuff to make finding this error sound simple but if exceptions were thrown instead of assigning NaNs to some variable life would be easier. The exception would tell me the exact line of the illegal operation while a NaN tells me nothing.
Javascript treats the "NaN" type as if it's a number... in the sense that when you do a math operation on something that is a NaN you get another NaN... The reality is... the very abbreviation "NaN" stands for is "Not a Number" so the addition of a number and something that is not a number should not even occur. Seriously? They should have called it "LoL" or "WtF" because these are more appropriate abbreviations.
I'm imagining that. It's my first day at a new job cleaning up a legacy codebase, but upon opening up a single 10kloc function written in JS, I realize that I made a terrible mistake and immediately hand in my resignation.
In the meantime, I set a breakpoint at the end of the function, inspect the inline value annotations for NaNs, and trace them back to the start.
Relative to any other debugging task in a 10kloc function, that doesn't sound bad at all. If it's consistently returning NaN, that should take maybe half an hour.
Also, I'd be really surprised if the function was made to return an int, because JS doesn't have integer datatypes.
> Sure, I can say all kinds of stuff to make finding this error sound simple but if exceptions were thrown instead of assigning NaNs to some variable life would be easier.
I said this in the parent post and sure enough the reply is exactly inline with what I said was going to happen. I don't hate javascript, but it has its horrible horrible warts.
>Relative to any other debugging task in a 10kloc function, that doesn't sound bad at all. If it's consistently returning NaN, that should take maybe half an hour.
An exception would take a second. An exception would give you the target line of every illegal operation; while a NaN propagates itself like a virus along every arithmetic operation.
Some people find it easy to run a 20 miles, sure but that still doesn't change the fact that cars are much better. Get the analogy? You are running 20 miles and telling me how you don't mind... I'm saying sure, but driving still has a purpose in life.
Breakpoints are off the table? Seriously? I didn't realizing using a debugger made me a JS fanboy.
Wrapping all of your basic arithmetic in try/catch blocks is completely unreasonable. If you need to check for NaN, check for NaN. Making arithmetic substantially slower in the general case to catch the uncommon cases is exactly the problem NaN was introduced to solve.
I'm not even saying to wrap your functions in try catches. If your program has a 1/0 operation then obviously you made a mistake somewhere. With javascript you have to hunt that mistake down, with a language like python you get the line number of where that mistake occurred. The exception is there for the entire program to fail so that you can correct your mistake. The language should not allow operations to occur that create these things called NaNs and thus automatically this removes the need to check for NaNs too. That's all I'm saying.
BTW unless you explicitly instantiated a NaN you should rarely ever be checking for NaN's in your code. NaNs represent illegal operations like 1/0, and thus any type of code producing a NaN should be killed. If you check for NaNs in your code that means you are guarding your code from your own bugs.
>I didn't realizing using a debugger made me a JS fanboy.
Yeah that was inappropriate. I apologize for calling you that, unfortunately you still saw it before I edited it out.
JS uses double-precision floating point numbers, based on the IEEE 754 standard first published in 1985.
[1]: Though there is a case that using doubles for all numbers is poor design, that's a different topic altogether.
Not to mention a null. null + 1 = 1. This is even worse. Your 1000loc function returns a legit number but it's a little bit off... and you have to find it. Imagine never having to have to do that with a language like python... 1+None and 1/0 all return named exceptions with line numbers in python.
Keep in mind this guy didn't even mention how javascript handles nulls. While nulls bypass typecheckers, typically a null + 1 throws an exception. Not so in javascript... null + 1 returns a 1 in javascript.
[1]: So far as I'm aware, though I'd love to read about it if I'm wrong.
Because all modern processors and mainstream languages implement their floating point calculations as IEEE 754.
> when did i say I want to make arithmetic slower?
When you said you wanted to introduce exceptions whenever NaN is encountered.
Your processor has instructions that add/subtract/divide/multiply floating point numbers according to the IEEE 754 standard. What you propose is to then check in the implementation of JS after each instruction to see if the result is NaN, which is going to be a speed reduction of at least 2-3x, though I would expect it to actually be higher than that because branching can get expensive. (Making a note to go try it and benchmark)
Then, after doing this check, you want to have JS throw an exception. This only slows down the uncommon case, so that's not much of an issue, but in situations where NaNs are able to be treated as valid values, they then have to catch these exceptions and resume normal execution flow.
The result is that there is a happy path slowdown of arithmetic operations by at least 2-3x, to gain an arguable advantage in debugging time in the unhappy path.
Throwing away compatibility with other languages and working against the processor are not goals I'd shoot for, and wrapping basic arithmetic in try/catch blocks doesn't sound like a very good payout for doing so.
Almost every popular language launches an exception when you do a divide by zero.
>Your processor has instructions that add/subtract/divide/multiply floating point numbers according to the IEEE 754 standard. What you propose is to then check in the implementation of JS after each instruction to see if the result is NaN, which is going to be a speed reduction of at least 2-3x, though I would expect it to actually be higher than that because branching can get expensive. (Making a note to go try it and benchmark)
Try it, I'd like to see the results. I'm not too familiar with the processor implementation, but something tells me that if every system level language implements exceptions for division by zero then the cost must be minimal. Also how would it be a 3x slow down? At most it's just one instruction.
Interestingly in some languages integer division by zero throws an exception and floating point division by zero returns some kind of infinite or NaN value.
If this behavior was propagated down from processor design, I'd say the processors made a bad choice. Why is this behavior specific to floating point? Why is the behavior for integers different. Either way when I say poorly designed, I mean poorly designed in terms of usability, not speed. Clearly, javascript wasn't initially designed for speed.
>Throwing away compatibility with other languages and working against the processor are not goals I'd shoot for, and wrapping basic arithmetic in try/catch blocks doesn't sound like a very good payout for doing so.
Like I said, the purpose of a division by zero exception is not to be caught. It's to prevent a buggy program from continuing to run. The only place a division by zero operation can legally occur is if the zero arrives via IO. In that case it's better to sanitize the input or detect the zero rather than catch an exception.
1.) Divide by zero returns Inf, not NaN.
2.) I just tried C++, Java, C#, Go[1], and the champion of safety: Rust. All of them evaluate to Infinity when dividing a double by 0.0. No exceptions to be found.
> Try it, I'd like to see the results. I'm not too familiar with the processor implementation, but something tells me that if every system level language implements exceptions for division by zero then the cost must be minimal.
But they don't. Not every system level language even has exceptions and the ones that do return Inf as stated above.
> At most it's just one instruction.
Branches can be expensive if the branch predictor chooses the wrong path and forces a pipeline flush. Should be able to work with the predictor so that happy paths are true and that should make it less likely, but I'm not an expert there.
> The only place a division by zero operation can legally occur is if the zero arrives via IO. In that case it's better to sanitize the input or detect the zero rather than catch an exception.
In many applications this is true. Not all.
[1]: You do need to make sure Go treats both operands as doubles, easiest way to be sure is to declare x and y as float64 and then divide them afterward. But when actually dividing a float64 by another float64, it does evaluate to Inf as well.
Well, Rust doesn't have exceptions, so it wouldn't happen, but yeah. To your parent, these languages follow IEEE 754, as far as I know, which specifies this behavior.
Ok overall I'm wrong.
But either way poor design from a usability standpoint and from a consistency standpoint because ints don't do this.
TIL about Go:
Numeric constants represent exact values of arbitrary precision and do not overflow. Consequently, there are no constants denoting the IEEE-754 negative zero, infinity, and not-a-number values.
Would probably retain it better if I used Go as my primary language for a year or two.
`Array.prototype.map` takes a function and calls that function for each element in the array, passing in the value, the index, and the array itself.
That means that your 0th element will be parsed as you would expect, your 1st element will be parsed as gibberish, your 2nd element will be parsed as base 2, your 6th element will be parsed as base 6, etc....
Now, should you treat the radix as optional? Absolutely not. I've got several commits on newly adopted codebases where I went through and explicitly used `10` as the radix anywhere `parseInt` was called. But is it technically optional? Absolutely.