I love that I have pretty much 0 experience with JS, and yet I can get every single question right by asking myself “what would be the craziest possible result here?”
The interview: "Invalid octal 018 that defaults to decimal minus valid octal 015" The job: "Should I try to add this through more polymorphism or slap a Visitor pattern into all these classes?"
For the Raw string question, in Python, we often use raw strings to type regex. They are already complicated enough, and we don't want to make them even more complicated by having to escape all the backslashes
Same main reason why String.raw was introduced in JS; to be used to create regex patterns as strings that can be compiled to RegEx objects later that can be processed. Before this you'd have to escape the escapes, e.g. /Hello World/ into `Hello\ \ World`.
@@RichardRemer The largest problem is the same issue I have with PHP not running in strict mode. Even if I'm coding in TypeScript, at runtime an API could return the wrong type! It's a philosophy decision. JavaScript and PHP are deliberately designed to continue execution with garbage data rather than fail.
I find that applies more to languages like rust where everything is “ugh” and “why” and “do I really have to do all this and write macros for something that would be 1 line in c# or js”. (Sorry, just learning rust and not having a great time. JavaScript looks pretty nice right now)
@@testacals oh I know that. But something like Swift is also lower level (arguably a little less than rust), but way easier to code in. A lot of the stuff I dislike about rust are design decisions
Hey man, just wanted to thank you. I'm learning to code and despite working in tech for over a decade, there's a part of me that has always resisted/dreaded coding because in my mind it has always seemed like the epitome of postmodern desk slavery. But you make it fun. You're always super hyped and having fun with coding. I'm not really learning that much from you, but listening to you having so much fun with it just makes it seem less soul sucking and has helped me have more positive associations with coding, which has really motivated me to do more learning in my spare time. Just a fellow anon or here appreciating your passion. Never change!
This is really interesting to me. Since in my mind code was always the thing that lets me make the computer do things. Like, it's the stuff that video games are made out of and the fact that I know how to do that now would probably make 5 year old me happy. (and also makes me happy now :) )
@@cameron7374 yeah I'm coming to see it more like this now, largely thanks to prime and LLL. It's weird - I had his experience with manufacturing first. When I started working with startups and got to see how things are made behind the scenes, actually going to suppliers, operating high tech machinery, working with vacuum chambers and plasma and lasers and CNC machines etc, it gave me a massive appreciation for the whole process. I was instantly hooked. But software always seemed like the drab side of things. Working with hardware was interesting and seemed like it had more room for creativity; coding just seemed laborious, just stitching together the hardware with boring 1s and 0s. It took actually taking a coding class and building some CLI apps in C to give me that same feeling about code. I see now how much of a rush it can be to problem solve with code, how intellectually stimulating it can be, and above all else how rewarding it is to build something that really works. I'm taking a break from studying at the moment because my daughter was just born a few months ago, but watching prime and keeping up my practice on sololearn has me excited to get back to my comp eng classes.
I don't know about Javascript, but in other languages, the main use cases I've seen for raw strings are: 1. Writing regexes without having to escape \ 2. Pasting the contents of a multi-line text file into a string literal. 3. Writing GLSL shaders or other dynamically compiled code from a different language, so they don't need to be shipped as separate files. 4. Writing multi-line strings for whatever other reason.
Pure rage.... But also, you construct new String objects with `new String()` and the fact that it doesn't crash when you do that without new is the part that's actually weird and wrong
@@itsteelworksof course. String is a function that converts things to primitive strings, and new String is a constructor that creates String instances. You can even do this with your own types, just check new.target in your functions or constructors and return a value to use instead! Just remember that you can't return a primitive if the caller used new, or it will ignore the value and return the this value it created for you. It's so simple and nice! 🫠
@@RichardRemer But why would a cast using String result in a string, I would not call that "consistent" at all. The sane expectation still would be that it casts to String, not string.
If you grew up with a typewriter (before eg graduating to a Vic20), the newline+return sequence makes perfect sense as they were two very distinct actions. 😂 Greatly enjoyed this episode.
I think the odd keyboard still has an enter/return key with an arrow on it that goes downwards and then points back to the left. The clue to 13 and 10 being two instructions hence two characters is even there.👍 But you're right. If you've used a typewriter before you can't fail to differentiate line feed from carriage return. I wonder... Should the 'return' key be called the line feed carriage return key for completeness? 🥴😂
I love that after 20min of bamboozling there was a phrase in an explanation to one of the questions: “if you do this instead you’d get a result you would have expected from the beginning”. You can’t expect anything after that
Just curious, does anyone here actually write conditions that depend on Javascript's type coercion system? Personally I avoid it like the plague and just make the most explicit conditions possible whenever I'm nose plugging my way through a JS build.
~26:40 - The rule is not completely consistent, at least not at first glance. Java uses a String pool for optimization and Strings that are referenced multiple times are therefore the same instance and `==` returns true. However that is not the case for all methods that can create new Strings. I do not know which methods create separate instances and which do not
It also differs based on JVM implementations, such as OpenJ9 having stricter string interning. You should definitely always use `.equals` when comparing strings, it short-circuit checks refs in the equals implementation anyway!
From the ECMAScript spec: 7.2.14 IsLooselyEqual ( x, y ) The abstract operation IsLooselyEqual takes arguments x (an ECMAScript language value) and y (an ECMAScript language value) and returns either a normal completion containing a Boolean or a throw completion. It provides the semantics for the == operator. It performs the following steps when called: 1. If Type(x) is Type(y), then a. Return IsStrictlyEqual(x, y). 2. If x is null and y is undefined, return true. 3. If x is undefined and y is null, return true. 4. NOTE: This step is replaced in section B.3.6.2. 5. If x is a Number and y is a String, return ! IsLooselyEqual(x, ! ToNumber(y)). 6. If x is a String and y is a Number, return ! IsLooselyEqual(! ToNumber(x), y). 7. If x is a BigInt and y is a String, then a. Let n be StringToBigInt(y). b. If n is undefined, return false. c. Return ! IsLooselyEqual(x, n). 8. If x is a String and y is a BigInt, return ! IsLooselyEqual(y, x). 9. If x is a Boolean, return ! IsLooselyEqual(! ToNumber(x), y). 10. If y is a Boolean, return ! IsLooselyEqual(x, ! ToNumber(y)). 11. If x is either a String, a Number, a BigInt, or a Symbol and y is an Object, return ! IsLooselyEqual(x, ? ToPrimitive(y)). 12. If x is an Object and y is either a String, a Number, a BigInt, or a Symbol, return ! IsLooselyEqual(? ToPrimitive(x), y). 13. If x is a BigInt and y is a Number, or if x is a Number and y is a BigInt, then a. If x is not finite or y is not finite, return false. b. If ℝ(x) = ℝ(y), return true; otherwise return false. 14. Return false. ℝ(x) stands for the mathematical value of x, which normalizes +0 and -0 into 0 and is not defined for non finite values.
The best part is JavaScript added “use strict” and none of these are addressed by this flag, you would think that octals would have required the 0o syntax. Maybe they should add a “really use strict” flag!?
@@fredoverflow You’re right, but how many people use “use strict” because of the limited improvements that come with it? I tried using “use strict” in REPL mode (both console and Node) and that doesn’t work, I should have written it in a file to test. It should have made coercion apply to only one side of the comparison and only be applied once, arrays should have required to be the same type for all elements, and add boxing of values. With the number of people who create videos like this, it shows that removing these problematic behaviors would lead to less errors for developers.
@@RichardRemer My problem with “use strict” is that it doesn’t apply enough changes to fix these tricky questions in the quiz. One of the biggest issues is inconsistency with type coercion, when sorting it always casts to string but then in comparisons it doesn’t and it can cast more than once. There needs to be a stricter set of rules that makes type coercion only work for comparisons and arrays are only allowed one type, this would fix the majority of the problems and have types that are predictable.
I wish you would put the links to the materials you are reacting to in the description, that seems respectful but also helpful, this has bothered me several times is when I wanted to try to go look at it myself. Love the content though. I wish I had the video that I made you react to once which was dave portnoy the barstools sports founder ranting about bitcoin, comparing it to "that emoji game" and so on, it was taken down for some reason. by the way, you are wrong about the conversion, 1 == "1" converts those to numbers, hence e.g. 1 == "+1.0e0", edit: lol sorry this was explained just a few minutes later, I think the explanation is that it's kind of like a special case that strings when compared to numbers is done that way, and then false is simply always converted into zero, but yeah I can't explain it either. it doesn't really make any fucking sense whatsoever
Mdn doc on '==' say It works ona case by case basis so: > Number to String: convert the string to a number... So the explanation by the quiz was correct.
@14:15 theprimeagen: *concentrating very hard to follow the exact logic behind Javascript's blackmagik type conversion* chat: uh-huh-huhhuh...he said "bang array" 😂
24:26 YES, I have used the String constructor for an actual functioning purpose. When you use the strings as map keys and two of those keys might be identical strings BUT you want to process the identical strings differently (for instance, based on their position or based on a user-selected process), then if you map the strings to their processing outputs, using String constructors lets you map two identical strings to two different processing outputs. If you didn't use the constructor, the second string processed would overwrite the first when you map that process output to the string, which you don't want if your goal is to process the two strings differently. The catch is that you have to handle all keys symbolically or retrieve them with a keys() function call if you want them, which is fine since you're probably going to be processing dynamic instead of literal strings anyway. Now, here's a bonus JavaScript quirk in the form of a quiz problem: let a = new String('hey'); let b = new String('hey'); let c = 'hey'; console.log(a == c); // guess the output of this first, then try it console.log(b == c); // guess the output of this second, then try it console.log(a == b); // guess the output of this third, then try it
I actually got wrecked by sort implicitly casting to string once. Worst part was that I wasn't actually the one writing the frontend I wrote the backend that consumed the sorted array.
tbf if you expect a sorted array and the client does not send a sorted array, that's an FE issue not a BE issue. But also there probably shouldn't be the need for a sorted array and the BE should sort the array itself regardless
@@xelspeththis. Everything should be done in the backend except fetch requests, storing cookies/temporary identifier variables, and DOM manip unless you have a specific edge case (such as using WASM or somehow not having a backend for your calculator website). If your company doesn't have control over the full tech stack though, then you're gonna have to contact the frontend devs to redesign your protocols for what data to send over the wire.
21:50 In the case that you're saying you don't know how to use it because you don't understand it, since it's raw but not really because it still evaluates the variables, the best way I can explain it is that that's only due to the backticks (``) surrounding the text, which I assume tells Javascript to evaluate the variables, and String.raw doesn't really interfere with this process. If it had the single or double quotes, it would have included the ${varOne} and ${varTwo}. And in the other case where you are wondering why you would need to use String.raw at all, I have found myself needing to use this at times when I was trying to either be able to log out the raw thing in LLM responses, since I was trying to figure out if it was giving or alone. I bet there's some dynamic finetuning data generation use cases too.
Very interesting video, I got almost all of them wrong, except for the String.raw I've had to use String.raw before when parsing JSON where some of the object attributes are also JSON (doubly stringified!), to avoid the backslashes in the double stringified json being used as escape characters
21:55 logging, debugging... when you want to see what's in the string. If whitespace gets printed regularly, it's difficult to see which type of whitespace it is.
3:10 the exact same question was in my midterm a couple of days ago and my smartass for some reason thought that typeof would return a sort of Wrapper class so like actually Number not a string (Java reference)...but that turned out incorrect obviously...
At 6:50 ish -- As far as I can find, even though it's counter-intuitive, I believe it actually tries to convert the string argument to a number to do the comparison. If you try 5 == '5.00000' it returns true. If it was as simple as the number argument being converted into a string, these would not be loosely equal. Instead '5.00000' becomes 5 *Edit* : Should've watched further, didn't think he'd come back to it. Prime talks about it again around the 9 minute mark
so the only time I've used a raw string was to identify different dating formats or directories with regex. but that was in python, idk if it would have the same application in JS
In the question 4, 0 == '0', '0' is converted to a number. The specs: Abstract Equality Comparison (==) The comparison `x == y,` where x and y are values, produces true or false. Such a comparison is performed as follows: 1. If `Type(x)` is the same as `Type(y)` then 1. 1. Return the result of performing Strict Equality comparison x === y. 2. If x is null and y is undefined, return true. 3. If x is undefined and y is null, return true. 4. If `Type(x)` is Number and `Type(y)` is String, return the result of comparison `x == ToNumber(y)` 5. If `Type(x)` is String and `Type(y)` is Number, return the result of comparison `ToNumber(x) == y` 6. If `Type(x)` is Boolean, return the result of the comparison `ToNumber(x) == y` 7. If `Type(y)` is Boolean, return the result of the comparison `x == ToNumber(y)` 8. If `Type(x)` is either String, Number or Symbol and `Type(y)` is Object, return the result of the comparison `x == ToPrimitive(y)`. 9. If `Type(x)` is Object and `Type(y)` is either String, Number or Symbol, return the result of the comparison `ToPrimitive(x) == y`. 10. Return false.
9:04 I incorrectly guessed it would be true, just because of the loose equivalence operator instead of a strict one. In my head I figured js would go the extra mile and either round to the same floating point placement or just chop off. But I was wrong lol
we had something similar as an exam in school. the results were literally that 25% of answers were correct. Which means the results of the entire class were the same that you would expect if everybody just answered randomly.
This is why I love the ThePrimagen: he turned an 11 minutes video into a 30 minute reaction. He's not just sitting there in the corner; he actually adds to the content. Love it.
26:25 BTW, JAVA handles strings with a stringPool. Each unique string has one reference in memory, so comparing the reference or the actual value is the same. Unless you compare “a” == new String(“a”), then the reference is different. But “a” == “a” is true
The only reason to use a string constructor is if you've monkeypatched or extended String with additional methods, which you shouldnt do 99% of the time
Apropos \f (also known as ^L or "form feed"): fun fact, GNU source code is partitioned with form feeds. I haven't see anyone else do that, but it's extremely common in GNU projects.
24:23 .. Yes! Exactly xD I was so offended just by thinking about how stupid and useless it is :D It's almost like `typeof null -> "object"` (but that is more like a bug) .. it's like and appendix of JS lang, but the appendix at least have some real function they say .. this is just worse than primitive string in every way except you may use it to some really bad programming like to make string object that inherite from another object or so, but even that bad programming could be done better without using new String.
'A very confusing conversion system' is what you tell someone just starting to learn js to give them a heads up without spoiling how atrocious everything is.
Java will not duplicate strings in the heap when created with double quotes, so "foo" == "foo" will return true even though it is technically comparing references instead of values. You have to explicitly create additional objects with new String() to break the equality.
I don't know who wants to know this info, but if you do arr.map(_ => 99) while the arr = new Array(5) --> means empty array of five. You can use fill before the map to turn all items to undefined then map through it. arr.fill().map(_ => 99) will work.
In Python, you use raw strings for RegEx or SQL queries to remove the need to escape everything. So for Python it is as easy as just putting an r in front of the string literal like this r’/t’, but JavaScript has so much more to type that it isn’t reducing the amount you need to type and I would argue that it isn’t making it that much easier to read.
There are regex literals for the first case :p But you are right, there actually are custom JS regex implementations as well as sql libraries that use this "tagged string literal" syntax - and the tag used for them is custom and typically shorter (like @shaunpatrick8345 showed, it can be any name in scope). It can be used in any context where you want a custom language string parser or transformer, such as latex, markdown, htmlx, jsx, etc... A tag's implementation has access to and can chose to use both the normal string version as well as the "raw" version in case it doesn't want the backslash escapes to have their regular string meaning. It also has access to interpolated values (parts that are in ${}) so it can escape them as it wants or even bind them as sql statement parameters or use them as jsx attrs or what have you.
The raw string stuff is used when using regex in a language using strings. Everything needs to be backslashed to count in regex, so it need to stay and not "dissapear" as it would in a real string. I did it in python tho so I dont know about javascript, but I assume its the same
8:42 It would've been fun if the question was baity and said. 0.2+0.3==0.5. _Although_ floating point can have inaccuracies, that's an example that works out.
The problem with Javascript and number is that there's only one representation and that representation is the IEEE754 floating point. Proper languages have decimal and integer types for use when you expect such semantics.
Wtf, I had to look it up because I didn't believe it, but Javascript really doesn't have integers. What a complete mess. How are you supposed to save integer numbers in the range above ~100000+? Do you have to create some patchwork integer class/struct? Do you have to use byte arrays manually?
@@brianviktor8212that's the neat part, you don't. Either use a different language for the backend or only store that value as a string and hope you never have to do maths with it. I mean, you could probably compile a custom NodeJS module that takes the string, parses it to an integer, does math, then spits it back out to JS as something that it won't mistreat as a float, but at this point it'd be easier to just use a different language for your backend
string is weird in JS. It is kept in stack and heap, but it is merged in heap by value. So, "some string" === "some string". However, if you use new String() it will create new entry in heap. Therefore, it will be false. JS is weird
One use case for String constructor is if you want to store other properties against the string (eg isTranslated) let s = "hello"; s.isTranslated = true console.log(s, s.isTranslated) s = new String("hello") s.isTranslated = true console.log(s, s.isTranslated)
I understood most of these, and I can accept most of the ones I got wrong, but the fact that string literals are not instances of strings despite their prototype being string makes me incredibly angry.
String.raw can be used to get syntax highlighting for the code inside a string, for example, make a function named "yaml" and just return String.raw from it, call the function somewhere, write some yaml inside the backticks and you get yaml syntax highlighting, at least if you use neovim with treesitter.
you'd use the constructor String in a case where you have a defined array of objects holding type/value pairs (i.e. a configuration map) and you want to map over it in your runtime and create the variables. here's a one liner: `const data=[{t:String,v:'my string'},{t:Number,v:12}];const instances = data.map(({t, v}) => new t(v));`
"5 empty items" becomes even more cursed because in some cases "empty items" from an array can end up in non-array objects with string keys, and then there is no way to remove them
I did some Performance testing because I have no hobbies, and I was generating giant tables of data with millions of elements. So doing new Array(1000000) reserves the memory cells in your ram to be used for this Array. So you have values that are next to each other. Same when you do some filtering with like while(len){array[len] = --len} it will allocate the values next to each other. If you do an array without a length then push to it, the array can have different adresses so like have gaps in it used by other variables.
I got I think 7/10 following along with this video even after studying JS since 1998. The octal one tripped me up because I didn't know offhand how JS would treat the unquoted value 018 given how relatively recent support for octals was finalized (and I don't actively use octals in my own JS anyways). Truthy-falsy ones only trip me up if I don't know which direction JS converts or coerces first when booleans are already involved because I made it a point to drill on that for hours based on this one particular article (from IIRC some Google/Opera dev like Hakon Wium-Lie or maybe jQuery creator John Resig or something) aggregating JS truthiness cases years ago (an article I'm trying to find, with the earliest search results giving me some Stack Overflow answers from 2013 or so instead) and I've literally never intentionally done "bool + bool" in my entire 25+ years of JS anyways. Finding out about empty array truthiness in this video was genuinely fascinating, BTW.
17:20 NaN2 would also make sense, number + string = 'NaN', 'NaN' + '2' = 'NaN2' But this is one of the few where JS actually makes sense as it treats everything equally
I love that I have pretty much 0 experience with JS, and yet I can get every single question right by asking myself “what would be the craziest possible result here?”
Yeah they should've thrown some normal curve ball questions, so people don't automatically default to answer with the craziest possible choices.
Except they did.. there were some normal expected ones there.@@kiattim2100
Just like in high-school, hahha...
they were just overcomplicating them. For most of them I could just look at it and be like yeah, 0 is false is true. duh.
You ain't fooling me
Can’t tell if I’m being gaslit or if I have been saying NaN wrong all this time…. 😅
I mean, there are multiple ways of saying it.
“Nan”, “N A N” or “Not a number”.
Just like SQL.
“S Q L” or “Sequel”.
@@Greedsmith "Squeel"
“I am currently a FAANG developer at this moment” - this was foreshadowing
severely underrated comment
what happened where does he work now
javascript feels like when i try to interpret my dreams
And it works, but in mysterious ways.
Javascript was a mistake.
This joke made me crack a rib.
Profile picture checks out.
😂😂😂
I finally understand the rust people.
Its javascipt devs who got out of a toxic relationship and now they want to date the opposite person.
The borrow checker might be annoying, but have you ever programmed in Javascript? - Rust devs
And after dating their opposite, theyll balance out in the middle, Go.
Or they take another detour and go elixir for a year
It's more like they have pathological tendencies and can't avoid toxic relationships, they went from idealizing JS to idealizing Rust.
Nah most of them go TS. Once they dip into the crazy they're hooked for life.
Funnily enough, I really like both JavaScript and Rust.
They complete each other.
The interview: "Invalid octal 018 that defaults to decimal minus valid octal 015"
The job: "Should I try to add this through more polymorphism or slap a Visitor pattern into all these classes?"
Uncaught SyntaxError: Octal literals are not allowed in strict mode.
@@fredoverflow hello this is CORS and fuck your app and especially those octal craps, I don't like them
perfect profile pic for this comment btw
When you can go Visitor, always go Visitor
roll your own rtti
For the Raw string question, in Python, we often use raw strings to type regex. They are already complicated enough, and we don't want to make them even more complicated by having to escape all the backslashes
It's also much nicer in Python because you just have to prepend an 'r' and you're good
And in Python you have to prepend 'f' to get variable formatting inside the string. (which you can combine with 'r' to get that French strings 🥖)
*python2 flashbacks*
Sounds like a skill issue. Real programmers escape all the backslashes and charge their employers for the time spent doing it. ;)
Same main reason why String.raw was introduced in JS; to be used to create regex patterns as strings that can be compiled to RegEx objects later that can be processed. Before this you'd have to escape the escapes, e.g. /Hello
World/ into `Hello\
\
World`.
It is truly amazing how much of JavaScript you miss out on by not doing aggressively stupid things with it
100% this
I just gave this comment its 70th like. Sorry about that.
@@RichardRemer The largest problem is the same issue I have with PHP not running in strict mode. Even if I'm coding in TypeScript, at runtime an API could return the wrong type! It's a philosophy decision. JavaScript and PHP are deliberately designed to continue execution with garbage data rather than fail.
If your code does not look like
var myInt = new Array(0b17 % spaghetti*4);
... you are not really coding
Yeah I hated every second of this
In a javascript quiz the answer is always "why"
I find that applies more to languages like rust where everything is “ugh” and “why” and “do I really have to do all this and write macros for something that would be 1 line in c# or js”.
(Sorry, just learning rust and not having a great time. JavaScript looks pretty nice right now)
quiz is the answer. why is the question
@@svenmify rust is comparatively low level than c# or js. That's why.
@@testacals oh I know that. But something like Swift is also lower level (arguably a little less than rust), but way easier to code in.
A lot of the stuff I dislike about rust are design decisions
@@svenmify
Looks like Harrison Ford recruited yet another aspiring rust cul... developer.
lol he updated the title cause he left Netflix 😂
Hey man, just wanted to thank you.
I'm learning to code and despite working in tech for over a decade, there's a part of me that has always resisted/dreaded coding because in my mind it has always seemed like the epitome of postmodern desk slavery.
But you make it fun. You're always super hyped and having fun with coding. I'm not really learning that much from you, but listening to you having so much fun with it just makes it seem less soul sucking and has helped me have more positive associations with coding, which has really motivated me to do more learning in my spare time.
Just a fellow anon or here appreciating your passion. Never change!
This is really interesting to me.
Since in my mind code was always the thing that lets me make the computer do things.
Like, it's the stuff that video games are made out of and the fact that I know how to do that now would probably make 5 year old me happy. (and also makes me happy now :) )
@@cameron7374 yeah I'm coming to see it more like this now, largely thanks to prime and LLL.
It's weird - I had his experience with manufacturing first. When I started working with startups and got to see how things are made behind the scenes, actually going to suppliers, operating high tech machinery, working with vacuum chambers and plasma and lasers and CNC machines etc, it gave me a massive appreciation for the whole process. I was instantly hooked.
But software always seemed like the drab side of things. Working with hardware was interesting and seemed like it had more room for creativity; coding just seemed laborious, just stitching together the hardware with boring 1s and 0s.
It took actually taking a coding class and building some CLI apps in C to give me that same feeling about code. I see now how much of a rush it can be to problem solve with code, how intellectually stimulating it can be, and above all else how rewarding it is to build something that really works.
I'm taking a break from studying at the moment because my daughter was just born a few months ago, but watching prime and keeping up my practice on sololearn has me excited to get back to my comp eng classes.
I am pretty sure the title changed from “FAANG engineer reacts to ExFAANG engineer” right after the announcement on the main channel lol
Alternative title: "When you create a language over a weekend"
10 days actually
@@fredoverflow 🤓
I don't know about Javascript, but in other languages, the main use cases I've seen for raw strings are:
1. Writing regexes without having to escape \
2. Pasting the contents of a multi-line text file into a string literal.
3. Writing GLSL shaders or other dynamically compiled code from a different language, so they don't need to be shipped as separate files.
4. Writing multi-line strings for whatever other reason.
I love this one even more:
String('123') instanceof String
JS is truly a language of all time
Pure rage.... But also, you construct new String objects with `new String()` and the fact that it doesn't crash when you do that without new is the part that's actually weird and wrong
@@itsteelworksof course. String is a function that converts things to primitive strings, and new String is a constructor that creates String instances.
You can even do this with your own types, just check new.target in your functions or constructors and return a value to use instead! Just remember that you can't return a primitive if the caller used new, or it will ignore the value and return the this value it created for you.
It's so simple and nice! 🫠
But Object('123') instanceof String is true
same for new Object()
Javascript, I love it
@@RichardRemer But why would a cast using String result in a string, I would not call that "consistent" at all. The sane expectation still would be that it casts to String, not string.
It cares more about debugging and foundational understanding than pure DS&A.
I have to admit, against my better judgment, I like this quiz.
JS's type coercion is a crime against humanity. Most of these questions should just TypeError…
If you grew up with a typewriter (before eg graduating to a Vic20), the newline+return sequence makes perfect sense as they were two very distinct actions. 😂
Greatly enjoyed this episode.
Gotta send the printer back to the left side of the page after new line.
Super fun yea
I think the odd keyboard still has an enter/return key with an arrow on it that goes downwards and then points back to the left. The clue to 13 and 10 being two instructions hence two characters is even there.👍 But you're right. If you've used a typewriter before you can't fail to differentiate line feed from carriage return.
I wonder... Should the 'return' key be called the line feed carriage return key for completeness? 🥴😂
I love that after 20min of bamboozling there was a phrase in an explanation to one of the questions: “if you do this instead you’d get a result you would have expected from the beginning”. You can’t expect anything after that
I love that you would never actually run into these problems if you code properly
Jr devs to the rescue!
Hearing expert JavaScript developers talk about its type system is a prime example of Stockholm Syndrome.
""""" type system """""
I'll have to lay down after this
Just curious, does anyone here actually write conditions that depend on Javascript's type coercion system? Personally I avoid it like the plague and just make the most explicit conditions possible whenever I'm nose plugging my way through a JS build.
Not anyone that has ever loved anyone in their life, I'm sure.
~26:40 - The rule is not completely consistent, at least not at first glance. Java uses a String pool for optimization and Strings that are referenced multiple times are therefore the same instance and `==` returns true. However that is not the case for all methods that can create new Strings. I do not know which methods create separate instances and which do not
It also differs based on JVM implementations, such as OpenJ9 having stricter string interning. You should definitely always use `.equals` when comparing strings, it short-circuit checks refs in the equals implementation anyway!
lmao, that was a short-lived title.
From the ECMAScript spec:
7.2.14 IsLooselyEqual ( x, y )
The abstract operation IsLooselyEqual takes arguments x (an ECMAScript language value) and y (an ECMAScript language value) and returns either a normal completion containing a Boolean or a throw completion. It provides the semantics for the == operator. It performs the following steps when called:
1. If Type(x) is Type(y), then
a. Return IsStrictlyEqual(x, y).
2. If x is null and y is undefined, return true.
3. If x is undefined and y is null, return true.
4. NOTE: This step is replaced in section B.3.6.2.
5. If x is a Number and y is a String, return ! IsLooselyEqual(x, ! ToNumber(y)).
6. If x is a String and y is a Number, return ! IsLooselyEqual(! ToNumber(x), y).
7. If x is a BigInt and y is a String, then
a. Let n be StringToBigInt(y).
b. If n is undefined, return false.
c. Return ! IsLooselyEqual(x, n).
8. If x is a String and y is a BigInt, return ! IsLooselyEqual(y, x).
9. If x is a Boolean, return ! IsLooselyEqual(! ToNumber(x), y).
10. If y is a Boolean, return ! IsLooselyEqual(x, ! ToNumber(y)).
11. If x is either a String, a Number, a BigInt, or a Symbol and y is an Object, return ! IsLooselyEqual(x, ? ToPrimitive(y)).
12. If x is an Object and y is either a String, a Number, a BigInt, or a Symbol, return ! IsLooselyEqual(? ToPrimitive(x), y).
13. If x is a BigInt and y is a Number, or if x is a Number and y is a BigInt, then
a. If x is not finite or y is not finite, return false.
b. If ℝ(x) = ℝ(y), return true; otherwise return false.
14. Return false.
ℝ(x) stands for the mathematical value of x, which normalizes +0 and -0 into 0 and is not defined for non finite values.
4. NOTE: This step is replaced in section B.3.6.2.
EXCUSE ME?
HE CHANGED THE TITLE OMG
The best part is JavaScript added “use strict” and none of these are addressed by this flag, you would think that octals would have required the 0o syntax. Maybe they should add a “really use strict” flag!?
"For real this time"
May I introduce you to Perl's pragmas :P
Both 015 and 018 do throw errors in "strict mode" though?
@@fredoverflow You’re right, but how many people use “use strict” because of the limited improvements that come with it? I tried using “use strict” in REPL mode (both console and Node) and that doesn’t work, I should have written it in a file to test.
It should have made coercion apply to only one side of the comparison and only be applied once, arrays should have required to be the same type for all elements, and add boxing of values. With the number of people who create videos like this, it shows that removing these problematic behaviors would lead to less errors for developers.
@@RichardRemer My problem with “use strict” is that it doesn’t apply enough changes to fix these tricky questions in the quiz. One of the biggest issues is inconsistency with type coercion, when sorting it always casts to string but then in comparisons it doesn’t and it can cast more than once. There needs to be a stricter set of rules that makes type coercion only work for comparisons and arrays are only allowed one type, this would fix the majority of the problems and have types that are predictable.
I wish you would put the links to the materials you are reacting to in the description, that seems respectful but also helpful, this has bothered me several times is when I wanted to try to go look at it myself. Love the content though. I wish I had the video that I made you react to once which was dave portnoy the barstools sports founder ranting about bitcoin, comparing it to "that emoji game" and so on, it was taken down for some reason. by the way, you are wrong about the conversion, 1 == "1" converts those to numbers, hence e.g. 1 == "+1.0e0", edit: lol sorry this was explained just a few minutes later, I think the explanation is that it's kind of like a special case that strings when compared to numbers is done that way, and then false is simply always converted into zero, but yeah I can't explain it either. it doesn't really make any fucking sense whatsoever
i do almost always have it... if its not in there, then its a mistake more than anything else. total accident if there is no link
Mdn doc on '==' say It works ona case by case basis so:
> Number to String: convert the string to a number...
So the explanation by the quiz was correct.
@14:15
theprimeagen: *concentrating very hard to follow the exact logic behind Javascript's blackmagik type conversion*
chat: uh-huh-huhhuh...he said "bang array"
😂
Besides regex, raw strings can also be for windows paths.
24:26 YES, I have used the String constructor for an actual functioning purpose.
When you use the strings as map keys and two of those keys might be identical strings BUT you want to process the identical strings differently (for instance, based on their position or based on a user-selected process), then if you map the strings to their processing outputs, using String constructors lets you map two identical strings to two different processing outputs. If you didn't use the constructor, the second string processed would overwrite the first when you map that process output to the string, which you don't want if your goal is to process the two strings differently.
The catch is that you have to handle all keys symbolically or retrieve them with a keys() function call if you want them, which is fine since you're probably going to be processing dynamic instead of literal strings anyway.
Now, here's a bonus JavaScript quirk in the form of a quiz problem:
let a = new String('hey');
let b = new String('hey');
let c = 'hey';
console.log(a == c); // guess the output of this first, then try it
console.log(b == c); // guess the output of this second, then try it
console.log(a == b); // guess the output of this third, then try it
2:10 I just learned that NaN is "Not a Number"... after all these years
I actually got wrecked by sort implicitly casting to string once. Worst part was that I wasn't actually the one writing the frontend I wrote the backend that consumed the sorted array.
tbf if you expect a sorted array and the client does not send a sorted array, that's an FE issue not a BE issue.
But also there probably shouldn't be the need for a sorted array and the BE should sort the array itself regardless
@@xelspeththis. Everything should be done in the backend except fetch requests, storing cookies/temporary identifier variables, and DOM manip unless you have a specific edge case (such as using WASM or somehow not having a backend for your calculator website). If your company doesn't have control over the full tech stack though, then you're gonna have to contact the frontend devs to redesign your protocols for what data to send over the wire.
this is so good.. Please do more of this.. Your channel is crazy awesome
21:50 In the case that you're saying you don't know how to use it because you don't understand it, since it's raw but not really because it still evaluates the variables, the best way I can explain it is that that's only due to the backticks (``) surrounding the text, which I assume tells Javascript to evaluate the variables, and String.raw doesn't really interfere with this process. If it had the single or double quotes, it would have included the ${varOne} and ${varTwo}.
And in the other case where you are wondering why you would need to use String.raw at all, I have found myself needing to use this at times when I was trying to either be able to log out the raw thing in LLM responses, since I was trying to figure out if it was giving
or
alone. I bet there's some dynamic finetuning data generation use cases too.
Javascript makes C's undefined behaviour look sane
I'm going thru like a worst phase of my life! Both professionally ans personally.
Watching your videos makes my day. Thanks Prime.
"If it's true, I'm gonna commit sudoku". Wow that's some high level of devotion to the cause man! :D
It's kinda entertaining to see chat slowly become completely hysterical, ngl.
Very interesting video, I got almost all of them wrong, except for the String.raw
I've had to use String.raw before when parsing JSON where some of the object attributes are also JSON (doubly stringified!), to avoid the backslashes in the double stringified json being used as escape characters
21:55 logging, debugging... when you want to see what's in the string. If whitespace gets printed regularly, it's difficult to see which type of whitespace it is.
now it's an ex-faang developer watches ex-faang developer take javascript quiz
6:55 why does this covert string into number, instead of converting number into string for the comparison?
Need to update the title as there are two ex-FAANG-ers here now))
Updated title: “ExFAANG Engineer Watches ExFAANG Take JavaScript Quiz”
I love the title change. 😄
3:10 the exact same question was in my midterm a couple of days ago and my smartass for some reason thought that typeof would return a sort of Wrapper class so like actually Number not a string (Java reference)...but that turned out incorrect obviously...
10:59 after array[6] = 9, is the length of the array 7, or 4?
Is JavaScript designed to be robust or just to be worst/more confusing than C++?
I like how Prime says "ha ha, look at chat in disbelief" and continues to be in disbelief for the next question himself 😔
Non-FAANG Engineer Watches FAANG Engineer watching ExFAANG Engineer take JavaScript Quiz
At 6:50 ish -- As far as I can find, even though it's counter-intuitive, I believe it actually tries to convert the string argument to a number to do the comparison. If you try 5 == '5.00000' it returns true. If it was as simple as the number argument being converted into a string, these would not be loosely equal. Instead '5.00000' becomes 5
*Edit* : Should've watched further, didn't think he'd come back to it. Prime talks about it again around the 9 minute mark
In the future fang engineers will be the people designing cyborg vampire teeth
I am unreasonably perturbed by the fact that `[]` can be coerced into both `true` and `0` depending on the type chosen.
so the only time I've used a raw string was to identify different dating formats or directories with regex. but that was in python, idk if it would have the same application in JS
In the question 4, 0 == '0', '0' is converted to a number.
The specs:
Abstract Equality Comparison (==)
The comparison `x == y,` where x and y are values, produces true or false. Such a comparison is performed as follows:
1. If `Type(x)` is the same as `Type(y)` then
1. 1. Return the result of performing Strict Equality comparison x === y.
2. If x is null and y is undefined, return true.
3. If x is undefined and y is null, return true.
4. If `Type(x)` is Number and `Type(y)` is String, return the result of comparison `x == ToNumber(y)`
5. If `Type(x)` is String and `Type(y)` is Number, return the result of comparison `ToNumber(x) == y`
6. If `Type(x)` is Boolean, return the result of the comparison `ToNumber(x) == y`
7. If `Type(y)` is Boolean, return the result of the comparison `x == ToNumber(y)`
8. If `Type(x)` is either String, Number or Symbol and `Type(y)` is Object, return the result of the comparison `x == ToPrimitive(y)`.
9. If `Type(x)` is Object and `Type(y)` is either String, Number or Symbol, return the result of the comparison `ToPrimitive(x) == y`.
10. Return false.
9:04 I incorrectly guessed it would be true, just because of the loose equivalence operator instead of a strict one. In my head I figured js would go the extra mile and either round to the same floating point placement or just chop off. But I was wrong lol
we had something similar as an exam in school. the results were literally that 25% of answers were correct. Which means the results of the entire class were the same that you would expect if everybody just answered randomly.
This is why I love the ThePrimagen: he turned an 11 minutes video into a 30 minute reaction. He's not just sitting there in the corner; he actually adds to the content. Love it.
"I am currently a FAANG developer at this moment"
now it will have to be an ExFAANG
26:25 BTW, JAVA handles strings with a stringPool. Each unique string has one reference in memory, so comparing the reference or the actual value is the same. Unless you compare “a” == new String(“a”), then the reference is different.
But “a” == “a” is true
The only reason to use a string constructor is if you've monkeypatched or extended String with additional methods, which you shouldnt do 99% of the time
Apropos \f (also known as ^L or "form feed"): fun fact, GNU source code is partitioned with form feeds. I haven't see anyone else do that, but it's extremely common in GNU projects.
24:23 .. Yes! Exactly xD
I was so offended just by thinking about how stupid and useless it is :D
It's almost like `typeof null -> "object"` (but that is more like a bug) .. it's like and appendix of JS lang, but the appendix at least have some real function they say .. this is just worse than primitive string in every way except you may use it to some really bad programming like to make string object that inherite from another object or so, but even that bad programming could be done better without using new String.
'A very confusing conversion system' is what you tell someone just starting to learn js to give them a heads up without spoiling how atrocious everything is.
So does undefined get converted to a string or number? What about special undefined?
Java will not duplicate strings in the heap when created with double quotes, so "foo" == "foo" will return true even though it is technically comparing references instead of values. You have to explicitly create additional objects with new String() to break the equality.
I don't know who wants to know this info, but if you do
arr.map(_ => 99) while the arr = new Array(5) --> means empty array of five.
You can use fill before the map to turn all items to undefined then map through it.
arr.fill().map(_ => 99) will work.
I'm not sure does that sort hurt more because it is 33 or because I know I would have made the same implementation if I was making it.
Edit: ExFAANG Engineer Watches ExFAANG Take JavaScript Quiz
i came here just to comment that, hahahaa
When did he retire?
@@QTA-20 he didnt retire he left netflix last week and gone full time content creator hence the ExFAANG
@@arafays That's really cool. Thanks
Principle of most surprise lmao. Actually insane that js became so dominant.
In Python, you use raw strings for RegEx or SQL queries to remove the need to escape everything. So for Python it is as easy as just putting an r in front of the string literal like this r’/t’, but JavaScript has so much more to type that it isn’t reducing the amount you need to type and I would argue that it isn’t making it that much easier to read.
const r = String.raw
I can't see that causing any problems...
There are regex literals for the first case :p But you are right, there actually are custom JS regex implementations as well as sql libraries that use this "tagged string literal" syntax - and the tag used for them is custom and typically shorter (like @shaunpatrick8345 showed, it can be any name in scope). It can be used in any context where you want a custom language string parser or transformer, such as latex, markdown, htmlx, jsx, etc... A tag's implementation has access to and can chose to use both the normal string version as well as the "raw" version in case it doesn't want the backslash escapes to have their regular string meaning. It also has access to interpolated values (parts that are in ${}) so it can escape them as it wants or even bind them as sql statement parameters or use them as jsx attrs or what have you.
The raw string stuff is used when using regex in a language using strings. Everything needs to be backslashed to count in regex, so it need to stay and not "dissapear" as it would in a real string. I did it in python tho so I dont know about javascript, but I assume its the same
8:42 It would've been fun if the question was baity and said. 0.2+0.3==0.5. _Although_ floating point can have inaccuracies, that's an example that works out.
The problem with Javascript and number is that there's only one representation and that representation is the IEEE754 floating point. Proper languages have decimal and integer types for use when you expect such semantics.
Wtf, I had to look it up because I didn't believe it, but Javascript really doesn't have integers. What a complete mess. How are you supposed to save integer numbers in the range above ~100000+? Do you have to create some patchwork integer class/struct? Do you have to use byte arrays manually?
@@brianviktor8212 You can store like 15 digits accurately without decimal place, that's good enough. If you need more you can always just use BigInts
@@brianviktor8212Javascript has a primitive type called bigint.
@@brianviktor8212 IEEE754 double precision can store integers up to 9007199254740992 exactly.
@@brianviktor8212that's the neat part, you don't.
Either use a different language for the backend or only store that value as a string and hope you never have to do maths with it.
I mean, you could probably compile a custom NodeJS module that takes the string, parses it to an integer, does math, then spits it back out to JS as something that it won't mistreat as a float, but at this point it'd be easier to just use a different language for your backend
javascript is the definition of "just because you can doesn't mean you should"
string is weird in JS. It is kept in stack and heap, but it is merged in heap by value. So, "some string" === "some string". However, if you use new String() it will create new entry in heap. Therefore, it will be false. JS is weird
I don't think JS even really has the concept of a stack and a heap.
Data seems to just kinda exist and the variables it's in have some scope.
I mean, it exists. There is a garbage collector as well. I don't think that the whole language is a lie 👨
Windows users know all about raw dogging strings C:\\señor
One use case for String constructor is if you want to store other properties against the string (eg isTranslated)
let s = "hello";
s.isTranslated = true
console.log(s, s.isTranslated)
s = new String("hello")
s.isTranslated = true
console.log(s, s.isTranslated)
I understood most of these, and I can accept most of the ones I got wrong, but the fact that string literals are not instances of strings despite their prototype being string makes me incredibly angry.
String.raw can be used to get syntax highlighting for the code inside a string, for example, make a function named "yaml" and just return String.raw from it, call the function somewhere, write some yaml inside the backticks and you get yaml syntax highlighting, at least if you use neovim with treesitter.
God! This is the best video I watch since the beginning of 2024
delete javascript
at 12:30 there are a few other ways to create empty items. one that comes to mind is [,,,,,,,,,]
"new Array" in JS can take a long walk off a short pier
Being real, if you are in a situation where you need to consider those situations, your code is already f*ucked from at least 2 sprints ago 😂
you'd use the constructor String in a case where you have a defined array of objects holding type/value pairs (i.e. a configuration map) and you want to map over it in your runtime and create the variables.
here's a one liner: `const data=[{t:String,v:'my string'},{t:Number,v:12}];const instances = data.map(({t, v}) => new t(v));`
"I am currently a FAANG developer... as of this moment"
Bro was telling us the writing on the wall.
String.raw`blah
` in JS is basically the equivalence of r"blah
" in Python, Rust, and a bunch of other languages.
JS is responsible for a whole generation of programmers thinking this sort of shit is acceptable and you can see it in their code.
"5 empty items" becomes even more cursed because in some cases "empty items" from an array can end up in non-array objects with string keys, and then there is no way to remove them
well the title now is incorrect
I suppose that JS victims may use raw strings to genererate dynamic code that includes a call that prints out a new line, for instance.
String.raw could be used for typing directories ?? Right 🤔
I did some Performance testing because I have no hobbies, and I was generating giant tables of data with millions of elements. So doing new Array(1000000) reserves the memory cells in your ram to be used for this Array. So you have values that are next to each other. Same when you do some filtering with like while(len){array[len] = --len} it will allocate the values next to each other. If you do an array without a length then push to it, the array can have different adresses so like have gaps in it used by other variables.
I got I think 7/10 following along with this video even after studying JS since 1998. The octal one tripped me up because I didn't know offhand how JS would treat the unquoted value 018 given how relatively recent support for octals was finalized (and I don't actively use octals in my own JS anyways). Truthy-falsy ones only trip me up if I don't know which direction JS converts or coerces first when booleans are already involved because I made it a point to drill on that for hours based on this one particular article (from IIRC some Google/Opera dev like Hakon Wium-Lie or maybe jQuery creator John Resig or something) aggregating JS truthiness cases years ago (an article I'm trying to find, with the earliest search results giving me some Stack Overflow answers from 2013 or so instead) and I've literally never intentionally done "bool + bool" in my entire 25+ years of JS anyways.
Finding out about empty array truthiness in this video was genuinely fascinating, BTW.
17:20 NaN2 would also make sense, number + string = 'NaN', 'NaN' + '2' = 'NaN2'
But this is one of the few where JS actually makes sense as it treats everything equally