Modernization Hub

Modernization and Improvement
React, Props, State – Lecture 2 – CS50’s Mobile App Development with React Native

React, Props, State – Lecture 2 – CS50’s Mobile App Development with React Native


[MUSIC PLAYING] JORDAN HAYASHI: Hello, and
welcome for Lecture Two– React, Props, and State. So last week, we talked about a bunch
of different topics, one being ES6 and beyond and the syntax
that comes with each of those. We talked about closures,
the process by which a function can reference variables
declared in a parent function. We talked about IIFEs, Immediately
Invoked Function Expressions. We talked about using functions
as first-class citizens. We talked about the execution
stack and event loop, and how JavaScript actually
executes in browsers. We talked about callbacks, promises, and
async/await, all of the different ways to handle asynchronous actions. And last, we talked about this
and the way that this is bound. This week, we’re going to
start with classes, which is a syntax I was introduced in ES6. It simplifies the defining of complex
objects that have their own prototypes. And with that, you have
two different things. You have classes and instances. Classes is basically an abstract
thing that you can declare, basically saying, hey, by
the way, any of these objects that you create will have these
methods associated with them. Or they might have these things
attached to them that you can use. And then when you actually
turn an abstract class into an instance of that object,
that is called an instance. An example of that would be the
difference between the date, which is a function– it’s a class– or a new date. So if you do something like
const d equals new date, then now you have a date object itself. And so Date with a capital D
would be a class in that case, and the lowercase D
would be an instance. These instances have things attached
to them, like methods and properties, and the classes have
things like static methods. And so methods are
basically anything that’s a function that can be invoked
on any of the instances. You can think of that as a
function on the classes prototype. A static method, on the
other hand, is basically a method that doesn’t really care about
the particular instance of a class. Instead, it cares about all instances of
the class, so something like Date.now, where you don’t really care about
a specific instance of a date if you just want to get the time. Whereas something like
turning a date to a string– in this class, d.toString– in that case, you do care
about the particular date object that you’re working on. And so when you do
capital Date.now, that’s considered a static method,
since it’s attached to the class. And if you do something like date– lowercase d– dot toString, that really
matters which instance you’re attached to, and therefore, that’s a method. Lastly, we have properties,
which are like methods. But rather than being functions,
they’re actually just values, and they are associated with a
particular instance of a class. And so with classes come
a few different keywords. We have new, which you saw me type
over here, which is basically saying, hey, give me an instance
of this particular class. You invoke the class like an object, in
case you want to pass anything into it. And so say we want to do new
const d2 equals new date, and we actually want
to pass in some number. That gives us a new date
from a very long time ago. So d.toString. Oh, d2.toString. Since we passed in the number 1,
2, 3, 4, that’s basically saying, give me a date that’s 1,234 milliseconds
after date 0, which is back in 1969. So a constructor is
basically something that you define within a class that says,
hey, when you create a new class, invoke this method such that you
create a new instance of a class. So let’s actually practice
this a little bit. So is everybody here familiar with
a data structure called a set? So basically what a
set is, is it’s a list, a data structure that supports things
like add, delete, and inclusion, where you cannot have multiple
things of the same value. Or in other words, it’s
a list of unique values. And the methods that
it should support are add, which is basically
add to this list; delete, which is basically get
rid of something from this; or inclusion, which is saying, hey,
does this list have a particular value? And it should also have the
ability to get its size. And so down at the bottom of
the file, I defined a few tests that we’re going to run after we
implement this, such as line 7 here, we have const s gets a new
set with an array from 1 to 5, which is basically saying, give me a
new set with five values, 1, 2, 3, 4, and 5. We’re going to try to
do S.add1, and so we’re going to try to add 1
three times to the set. And when we do S.size, we want it
actually to only have five members, because you shouldn’t be
able to add 1 multiple times. Down here, we do S.add6,
and then we try S.has6, and it should contain the number 6. That thing should not be there. We try to see the size of it, and
it should have added another member. And then down here, we tried
to delete that and do a couple of associated checks with that. And so how are we going to go
about implementing this class? So as per the slide over here, we
use a method called constructor in order to go ahead and construct
an instance of this class. And so within the class, we should
have a method called constructor, which takes a single argument. It should take an array or some
sort of list to be created with. And what are we going to do? Well, we should attach to
the instance the array. And that’s basically saying,
when I’m constructed, I expect to have one argument,
which is just an array. And the only thing I’m going to
do when creating this set object is just store a reference to that array. By doing this.arr=arr. And so in this case, the
this keyword is referring to the instance of the object. Cool, so let’s try adding a couple
of different methods to this. So we should be able to support
add, which should take a value. We should be able to support
delete, which also takes a value. And we should have something called
has, which checks for inclusion. So how might we do something? How might we add to this class? Does anybody have any ideas? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Exactly. So we should use something
like push to add to the array. But before we do that, we should make
sure that that number does not already exist. And so maybe we should implement the
has method first, which is a great idea. Let’s go ahead and do that. So how might we do has? Well, it turns out on
the array prototype, we already have something
called includes, which tells us if an array includes a value,
so we can just do that. We can do return this.arr.includes(val). And so now that we have that,
how might we take care of add? We should say, oh, well, if this does
not have the value already, add to it. And so here I use this.has. So this here is referring
to the instance of the set, and so this.has is
referring to this method down here, on this particular instance. And then when I do this.arr, this still
refers to the instance of this set. And so we’re just getting at
this array property that we have, and we’re pushing that
value to that array. Cool. So how might we go about this delete? AUDIENCE: [INAUDIBLE] has value. JORDAN HAYASHI: Yeah, we can
check if we have the value, but it doesn’t really
matter all that much. A quick and easy way would just be
doing this.arr=this.arr.filter and then we can just
filter by the values. So we could say, oh, we
want for every x in here, we want the x’s that
don’t equal this value. Cool. And so we can go ahead and
run this to see if it works. Oops. While I edit, I should flip these. And we go ahead. S should have five members and actually
has– ooh, we forgot to influence size, actually. And so we took care
of all of the methods, but we didn’t include the size. And so we should be able to
return the size of this set. How might we do that? Well, this is interesting. So we should be able to get at this
value by doing the instance.size. And JavaScript actually
has a convenient way. We can do get size,
which is saying, when somebody tries to get at the
value or the property.size, actually run this function. So this is just syntax
for that shortcut. So how might we implement size? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah, just
return this.arr.length. So now we can run this, and we
see I should have five members, and it actually has five. So that’s good. S should contain 5. That’s true. That works. S should contain six. It’s true. S should have six members, and
actually has six, so we’re good still. S should no longer contain six. That also returns true. And lastly, S should have five members,
and actually does indeed have five. So does anybody have any questions with
our implementation of set as a class? Great. So it turns out JavaScript
actually already has a set class, and it works exactly as we implemented. But say we actually wanted to use
the native implementation of set and actually add some stuff to it. So that’s where we use these other
keywords, called extends and super. So extends is the
JavaScript way of saying, hey, I want to start with a base
class and actually add to it. Extend this class. And super, as we’ll see in a second,
is when we’re writing that class, so we can refer to the original
class using this keyword. And so in this example
called my set, we’re going ahead and extending that set
with a bunch of different things. And so here you see constructor. It still takes an array. And the first thing that we do
is we invoke super on that array. So this is basically saying,
hey, we’re extending a set. So when you do the constructor,
the first thing you should do is actually run the
original set’s constructor. And then let’s also keep
track of this.originalarray is the array that’s passed in. And so we’ll use that later to
use this thing called reset. And so say we wanted to every single
time we added a value to this set, we also wanted to log, hey,
we added a value to this set. So we can just start writing
this method called add. It takes a value just
like any other example. But instead of implementing
add ourself, we’re just going to use the native
implementation of add. And so that’s where we use super.add. So super, again, refers to the
class that we’re extending. And so when we invoke
super.add, it goes ahead and does that using the
native implementation. And then since we’re going to extend
it with some additional logging. We’re just going to log, hey,
we added this val to the set. And if you’re not familiar with
this, if you use the backticks, you can go ahead and add
variables in line, in the string, and it’ll go ahead and
substitute those in. And so you see we added a
couple of other methods here. We have to array, which is basically
saying, hey, I actually want the array, and not the set. And so we can just
return Array.from (this). We’re passing in the entire instance. And lastly, we have a
reset, which is saying, hey, I want the original set
that I had, or at least a new set with equivalent value. So you can return a new my set. So notice you are referencing
my set inside that class. We want a new one, and
we’re going to pass in, as the array here, the original array. So this is an example of us extending
a class that already exists. And as you see, if we want to
reference methods on that original, we just use that super keyword. So any questions on sets,
how we define them– or sorry, classes and how we
define them, how we extend them? So why might this be useful? So as you guys are
doing on your project, you’re keeping track of
these things called to dos. What if we actually
had a class for to do? And when you invoke this constructor
on some configuration object, what if it pulls out the text
and whether it’s checked or not and stores it as part of
[? its ?] class instance? And say we want to
render it to the page. What if we could just
return some HTML like this? It would be quite handy, right? And that actually is
our next topic, react. So react is a JavaScript
library, and it allows us to write declarative views that will
react to changes in data automatically. It allows us to abstract complex
problems into smaller components, and it allows us to write simple
code that still perform it. And so I use this word in the
first bullet, declarative. So what the heck does that mean? So in CS50, we learned a paradigm
of coding called imperative. And today we’re going to talk
about declarative coding. So the difference in
imperative and declarative is like asking the difference
between how you do something and actually what you want out of it. So imperative programming
outlines a series of steps to get to what you want,
whereas declarative, you just say what you want. And it’s just an implementation
detail on how to get it. And so we’ve learned a few different
languages through CS50 in this course. A couple that come to mind
are HTML and JavaScript. So in HTML, do we tell
the browser exactly how we want to render all of these things? Do we tell it exactly how we
want the DOM to be constructed? No, we just tell it what we want. And so HTML is considered a declarative
language, because you just say, hey, I want this. And browsers are in charge of
just giving you what you want. Rather, with JavaScript, as you’ll
see in your first project, when you want to do anything to
the DOM with JavaScript, you tell it, oh, first
get me a new element. Call it a div. Then do this. Then maybe append it to the tree. Then maybe add a class to it. Maybe give it some inner HTML. And so you’re telling it exactly
what you want and how to do it. And so that is the more
imperative way of programming. So let’s take this into an example. Say we had a classical
guitar here, and say we wanted to actually create this guitar. So in an imperative way,
how would you describe that? Well, you would say, oh,
I need a head over here. I need to add some pegs to it. Maybe I want the neck. Maybe I add some frets to that. Oh, I need to create the body and attach
them all and then maybe return that. And what would be a more declarative
way of creating the guitar? You just say, I want a guitar. Maybe tune the strings to this. And so an example in pseudo
code would be like this. So say we have a guitar, and
say we have some function called create element, similar to
what we have in the document in HTML. And say we know exactly
what strings we want. How might we go about
creating this guitar? Well, first we might want to do
something like let’s create a head. Again, telling whoever’s
listening exactly what we want. And then maybe for 6 pegs, maybe we
want to start adding pegs to that head. And so now we, in a very terse
manner, have a head with six pegs. And so we might do the
exact same thing with neck. And maybe we want to add
something like 19 frets. And you can see how this gets a little
bit annoying to write it, right? And maybe we do the
same thing with body, but we have some sort of for
each loop with the strings, and we pass in the tone that we want. And then we go ahead and create a
string and tune it to that tone. And then we go ahead and
add that to the body. And you see how this gets very
step-by-step, exactly what we want to do to create a guitar. And this is exactly what we’ve been
doing thus far for writing to the DOM. And so how might we do this
in a more declarative manner? Well, we would just
say, give me a guitar. Give me a string. Maybe I want it to be tuned to
the first note in the string. And maybe copy that a few times. And there we go. A better way to do this would be
rather than hard coding these, maybe we just do strings.map, and
for each note, we stick it in there. So it looks like there’s
a little bug here. I used string even though I declared
the variable called strings. So if we map over the
array called strings, and for each note return a
string where the note is note, then we have declaratively
written a guitar. Does this make sense? So a great thing about react
is that the way that you code is in a very, very declarative manner. The browser APIs aren’t
super fun to work with. You get to work with them a bit
in project zero, but react just allows us to write exactly what we want. And the library will actually take
care of the DOM manipulation for us. And so what does that really look like? So say we wanted to create a slide here. Say we wanted to use the
native DOM manipulation API in order to create the slide here. So we might have a slide
element that we created and say we wanted to add a H1 to that. Add a title to it. How might we do it? Well, we’d have to do something like
const title=document.createElement. Get an H1. And then start adding to that. We can do title.innerHTML
is equal to the SLIDE.title. And you see how this starts getting
to like what we’ve been doing earlier, where you say exactly what you want,
but it might take you a long time to do. So in react land, this
is actually a lot easier. So how might we do this if we were
doing this completely declaratively? Well, we just say exactly what we want. We want a slide. Maybe it has a title, where the
title is equal to the slide’s title. Maybe we have some bullets. Or we can just map through the
bullets that we have up there. We can do SLIDE.bullets.map. And we can say for every bullet,
just give me a list item. And maybe we should wrap those
list items in unordered list. And maybe instead of using
this, we can do an H1 here. So see how this is a
lot more declarative? Makes sense, right? It just makes sense. It’s easier to read, and
it’s easier to maintain. So another great thing about react
is it’s very easily componentized. What do I mean by componentized? Well, it’s a process by which you break
a very complex problem into a bunch of different, discrete components. You can reuse these components,
which is great for consistency. So say you had a bunch
of slides, and you wanted to just change how every title looked. Well, if you did this using
native DOM manipulation, that might be a bunch of different
lines of code that you have to change. But if you did this
using React components, it might just be one line
that you have to change, and then it’s applied to every
single slide that you have. It’s also great for iteration
speed, because then you can just reuse the
components over and over, rather than having to cut and
paste code all over the place. Another great thing
about these components is that React’s declarative nature makes
it very easy to customize components. So say we had the last three slides
and wanted to write them in HTML. It might look like
something like this, where you have a div, which represents
a slide where the titles react. You notice these are the same
three bullets as previously. We have the declarative slide, where
we have the same bullets as previously. And say we wanted to change this. Say we wanted to make all of the
titles have a slightly different style. Well, how might we do that? Well, we’d have to edit this line. We’d have to go edit this line. Maybe down there, there’s another
title that we have to change. Or what if we wanted to even change the
structure of how we represent bullets? Maybe rather than using
an unordered list, we might use an unordered list with
a div that wraps the list items. Well, in order to do that, that would
be a lot of code you have to change. But imagine that we had broken this up
into a bunch of different components. Where might it make
sense to break it up? Well, you see us repeating code
that looks pretty much the same over here, right? So maybe that should be broken
up into a separate component. So how might we go about doing that? Well, first we have to extract the
information, and so in this array here, we have the slides where
I’ve basically just ripped out the information for each slide. And so now let’s go ahead
and implement that slideshow. And so in order to do
that, we might have something like where we have a
slide, which might take an array, or it might take an object. Let’s call it a slide. And what will it return? Well, it should return a div. Maybe inside it we have that H1. Then maybe we have that unordered list. And then go ahead and close that div. So what goes in the H1 here? Well, maybe we should
have that slide’s title. And what about generating those bullets? Well, like we did in
the previous example, we might want to do the
slide.bullets and map [? overflows ?] to create those list items. So for each bullet, just go
ahead and create a list item where you have that bullet. And so now we declared
exactly what a slide is. Well, a slide takes in
an object representing the slide in the same shape
of the object up here. And then it just returns the same
HTML that we had in the other example. It’s a div that wraps
in H1, where we stick in the title, an unordered
list that has some list items that represent those bullets. And now how might we
create that slideshow? Well, maybe we just do for
each slot in the slides, maybe we have an outer-wrapping
div that maps through the slides. And for each slide, we actually
just use the slide down there. So now we’re going to get ahead
and use React and generate basically the same HTML as
we had in the previous one. But now say we wanted to change
the styling of the title. How would we do that? Well, that’s just right here. What about if we wanted to
change the structure of how we created those bullets? In the previous example, if we wanted
to wrap the list items in a div, it would be many, many lines of code. But here, it’s just
you stick a div here, because this is basically an
abstract component that says, if we wanted to create a slide,
here’s exactly how we do it. Just make sure to pass me a
slide of the correct shape. And so the React paradigm is
to take a very complex problem and break it down into
small chunks like this that each solve a discrete
problem in that UI. Any questions so far? Cool. So another great thing is
that React is very performant. We just write what we
want, and React will do the hard work of
playing with the DOM, or changing things so
that what we have written matches with what we
have showing on the page. And so the way it does this is through
an algorithm called reconciliation. And this is the process by which
React syncs any changes in app state to the DOM. So what’s great about
React is that it actually maintains what’s called a virtual DOM. And so the slow thing
about playing with the DOM is that anytime you destroy elements
or create new elements, that takes a relatively long amount of time. And so the way that React
says, maybe we should do this better by rather than
changing stuff in the DOM every time we want to change something,
what if we just stored everything in memory? And then with any
changes, we can just say, hey, how does this differ
from what’s shown on the page? And only change what’s necessary. And so that’s what’s done in
the reconciliation process. So first, anytime your
data changes, React will reconstruct that virtual DOM. Then it will diff that DOM against
what’s actually there in the real DOM. And it’ll make only the
changes that are needed. Of course, there is an asterisk
that goes along with that, but in your mental model,
you can just think of it as React will just only make
the changes as necessary. So how the heck do we write React? Well, there’s this
thing called JSX, which is short for JavaScript and XML, which
is basically just an XML-like syntax extension of JavaScript. It’s great because it just
transpiles directly to JavaScript. So as we talked about
in an earlier lecture, the paradigm now is to rather than just
writing JavaScript by hand, using ES5 directly, we can write
in these other languages that just transpile back to JavaScript. So JSX is an extension to that
JavaScript syntax, which will just compile back down into JavaScript. And so tags defined in JSX, XML looks
exactly like HTML with the tags. Lowercase tags are treated
as HTML or SVG tags, whereas things uppercase are
treated as our custom components. And what the heck is a component? Well, a component is just a function,
a function that returns a node. What’s a node? It’s something that React can render. For example, like a div or a
span or a string or a number. And components will receive
an object of properties that are passed to this element. So going back to the example
that we were doing before, see how one, we declared a
const called slide down here. And notice how I
uppercased that S. That was fully intentional, because in
JSX, a tag with a capital S gets invoked like a function. And what does it get invoked with? Well, anything you
pass it as properties. So if you remember properties
from HTML, you just pass them in line with
the key and a value pair. In React, you do the exact
same thing, and it actually gets passed as an object to the
component that you described. And so now to talk about props. Well, props are just past as
an object through a component and used to compute the return node. So basically, the object that we were
just referring to in this example, we called it a slide. In React convention, we always call
that props, short for properties, like those HTML properties. And any change to these props will cause
a recomputation of that return node, or in React vocab, a re-render. And so unlike in HTML, these
can be any JavaScript value. So if you remember back to HTML,
when you declare a property, it has to be a string. It might be JavaScript,
like a JavaScript function, but really it’s just a string. And so you can’t pass any complex
objects or a date, object, classes, instances, and stuff like that. But in React, we can pass anything. So notice here, slide is an
object as declared in the array that we have up there. And we’re just passing
the object over there. And the way that we pass
values in JavaScript is we wrap them in single
curlies, which means, hey, execute this as a JavaScript. Does that makes sense? So let’s actually use
a live example of this. So there’s this website
called codesandbox.io, which is linked on the
website in the Resources tab, and this allows us to write React
and actually render it live. And so can everybody read this? So basically, what’s on the left
over there is code that gets executed and runs live on the right side here. So if we started editing this, you see
it applied directly on the right side there. And so you see we have
a const called app, which doesn’t care
about any arguments it receives and just renders
the static HTML here. So what if we actually
wanted to pass it a prop? So say we passed it something like
a count and passed it the number 1. Again, we use curly braces there to
represent this is JavaScript coming, and the JavaScript is just the number 1. So how might we go ahead
and render that in the app? Well, first, we need to actually
pay attention to the object that we receive here. And second, we can just
do it directly in line. And so we use those curly
braces to represent, hey, here comes some JavaScript. And we can just use
props.count, and now we’ll go ahead and grab the prop with
the key [? call ?] account. And so I mentioned that as props
change, React gets re-rendered. So how might we do that? Well, we can just wrap this in a set
timeout, or a set interval, rather. And every 1,000 milliseconds
or 1,000 milliseconds, let’s go ahead and pass it not
just one, but the count plus plus, where count starts at 0. And every time this is
invoked, it increments. And so down here, we’re
saying, sudden interval, which means run some
function every n seconds. We set n to equal 1,000
milliseconds, or one second there. And every time we want
to render this function called app with a count of one
more than it was last time. And so we start at 0,
and every second you see we pass a new prop
count incremented. And even though we’re not really doing
anything actively in the app here, we wrote this in a very
declarative nature, right? We said, for any props that you give
me, just render a div with an H2 where the H2 is the value of the props. See how that’s very declarative. We don’t say, oh, first
you have to grab the count, then check to see if it is
different than the last one, and then append this to the DOM. We just say, hey, this is what we want. And React will give it to us. Any questions on React
or props thus far? Yeah? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yes, of course. So set interval requires two arguments,
where the first one is a function. So the question was can you go
over the arrow notation here? And so set interval
expects two arguments. One is a function that it
will invoke with no arguments, and so I’m just declaring an anonymous
function here with no arguments. So I’m saying, here is a
function that takes no arguments, and what should I do? Well, I should render this. I could have also written this as
this, and it would have functionally been the exact same. Does that makes sense? Any other questions on React or props? Yeah? AUDIENCE: So in line 10 [INAUDIBLE]? JORDAN HAYASHI: So in line 10, we
have braces after the arrow here. So that’s just saying, interpret
all of this as one statement. So something special
about arrow notation is that it has an implicit return,
which means we don’t actually have to write– so we could have
written the same thing as this, as a function that takes props. And what does it do? It returns this. So this actually does
the exact same thing. So we’re saying, app
two is a function that takes a single argument called props. And what we do is we return this, and
so arrow notation shorthand is just you have your arguments in arrow. And then if you don’t have braces,
which referred to it like a code block, it just returns whatever’s next. And so we’re saying, return this. What is this? Well, it’s this div and H2. We’re going to wrap it in a parentheses
so we know that it’s just one value. Does that make sense? A great question. Do you guys see how these are the same? So this is great, but we don’t really
have all that much power yet, right? If we wanted to go ahead
and change these props, we still have to drop down to
this raw JavaScript over here. So next we’ll see exactly how we
create apps that are stateful. What does that mean? Well, there’s this
notion of state in React, and state is basically an
internally managed configuration for any component. And so now components become classes,
and this .state is a property on that component’s instance. So how do we update the state? Well, there’s a method
called this.setState, which is implemented in this thing
called a React.Component, which we have to extend in order to
have access to that method. And this goes ahead
and changes that value. So you can pass an object to be merged
or a function of the previous state. And so if we pass
this.setState, an object, it will go ahead and merge that
in with the existing state. So if we pass it in an
updater function, it’s basically a function that gets run
when we want to change the state. And the set state calls are
batched and run asynchronously. And of course, any change in
state will also cause a re-render, because it would be silly if
we were to change the state but not reflect that in the UI. And so how might we go about
representing state over here? So first, let me copy this
so that we can save it. So let’s go ahead, and rather
than having an app be a function, let’s actually have it be a class. So we can do class app, and we
want to extend React.Component. And within that, we want to have
this method called render, which is automatically invoked on a re-render. Within render, we want to return this. Cool. So the way to now write
this is rather than having app be a function that takes
props and returns something, we’re actually writing a class for app. And so as we talked about
earlier, classes have instances. And React knows that when
you want to render something like this, if it’s a class, go ahead
and create a new instance of that and pass in these as props. And notice how we don’t ever
take the props anywhere. That’s because when we
extend React Component, React Component, that base class,
goes ahead and attaches the props to the instance. And so in order to get at them, rather
than doing props does something, we do this dot props dot count. So again, the props that come in,
in the way that React.Component is implemented, it
automatically takes the props and attaches it to that
instance of the class. And so in order for us to get
them in the render method, we do this dot props dot count. Does that make sense so far,
going from a function to a class? We’ll talk about this in
depth the next lecture. And so we talked about
this thing called state, and how do we actually
go ahead and use that? Well, when we want to create
our state, we actually do that in the constructor method. And so the first thing that we want
to do in our constructor method is actually called a
super, which means allow React.Component to do
stuff with the props that it would have done otherwise. And now go ahead and
do what we want to do. What do we want to do? Well, we want to initialize
this thing called state. Cool. So now we have this thing
called state, and how are we going to go ahead and update it? Well, maybe we should have something
called increase count, which is a method on this instance. And let’s go ahead and
increase the count here. So how might I do that? Well, I should call this dot set state
and pass in count is this dot state dot count plus 1. And so now we have a method on
this instance called increase count that we can call. And it should, in theory,
increase that count. And so rather than referencing the
props down here, let’s reference state. And now it should be
0, and it’s just going to stay at 0, because
we’re not doing anything. So we can go ahead and
get rid of this interval. And so how are we going to change that
value of 0 to be something like 1? We have to invoke that
method, and so let’s go ahead and do that when we click a button. So let’s have something like
div here, and within that div, let’s create a button. And when we click that button,
let’s go ahead and invoke this thing called increase count. And let’s call it increase. So now if we click this button,
what do we expect to happen? So notice how we pass
this thing called unclick. What’s the value of unclick? Or what function should we invoke? Well, we should invoke this dot
increase count, which should be this. We’ll see what it is in a second. And then that should call
this dot set state, which should, in theory, increase count. So what happens? Uh-oh! Cannot read property
set state of undefined. So there’s a bug here. Can anybody spot the bug? Cannot read property
set state of undefined. It even tells us exactly where that was. So we’re seeing this
dot set state, which I told you is a method that we can use. But it’s saying, we cannot
read a property of undefined, which implies that
this is undefined here. Yeah? What do you think is going on here? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah. So we have to bind
that function to this. So we talked about this and binding
this towards the end of last lecture. And we talked about how this is bound
on whatever object it’s invoked on. And so right now, this isn’t
invoked on the class, right? It’s invoked when you click that button. And so we’re going to have to bind this
to the value that we want it to be. And so towards the end of the lecture,
we talked about different ways to bind this. Can anybody remember those? Yeah? AUDIENCE: Bind method. JORDAN HAYASHI: Yeah. There’s this thing called bind,
which is a method on all functions. And similar to bind
there is call and apply. And we also talked about a
different way of writing functions, which actually lexically binds this. Do you guys remember how that was? Remember, if you use arrow notation,
it automatically binds this for you. And so you can solve this problem
in all three different ways. And so down here, if we
wanted to do dot bind, we could do dot bind to this,
which, since we’re invoking it on the render method, which is a
method that’s invoked on the class, this is bound correctly. And so when we increase
it, it will increase. We could also use– well, call and apply
don’t really work here. But we could actually
use arrow notation. So we can say, let’s have
an arrow notation down here, which creates a new, anonymous function. And what does that
anonymous function do? Well, all it does is
invoke increase count. And so since this we wrote
as an ES6 arrow notation, it automatically binds this
to be what we want it to be. Does that makes sense? There are a couple other
places that you could do this. We could also do it in
the constructor method. You could do this dot increase count
equals this dot increase count dot bind this. Or lastly, you could also– and then down here, you’d just
have to do this dot increase count. Or alternatively, you could actually
just define this in the constructor. And so that would be the other way. So you could say, this dot increase
count equals an arrow notation in this. But let’s stick with this way for now. Cool. So now we have a button that works,
that increases as we expected. So what if instead of increasing it
by 1, we wanted to increase it by 2? Well, of course, you could just do this. Let’s say we want to do
it by sending state twice. So what do we expect to happen here? We expect it to set state to be count
as this dot state dot count plus 1 is 1. And then [? it does ?] set count
to be 1 plus 1 to be 2, right? But we see it’s only going up by 1. Why might that be? Well, set state calls are actually
batched and run asynchronously, which means React is smart enough to
know that if this dot set state is getting called a bunch
of times in a row, rather than immediately doing that,
maybe it would be better to just batch them and do it all in one go. And like I said, if you pass an
object to this dot set state, it will just merge it
into the new state. And so what is actually
happening when you call this dot set state twice in a row? Well, they get batched together. And so it says, OK, now
we know we need to merge into the old state these two objects. And so it’s basically saying, merge
the current state– in this case, it’s count is 0. And then what are we merging into it? Well, we want to merge this
dot state dot count plus 1. And then we also want to
merge that same thing again. And so what happens when you
merge these three things together? Well, these two get merged together,
and it actually turns into this. Well, you go left to right,
so count gets merged, gets replaced by this
dot state dot count, which then gets replaced by
this dot state dot count plus 1. And so that is actually just 0 plus
1, which is why we end up getting a 1 here. Does that make sense? And so we can prove that it
runs asynchronously by logging. So what do we expect this dot
state dot count to be at line 21? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: It actually logs 0,
because when you call this dot set state, all it does is
add it to that queue that we know it’s
eventually going to execute. Same with this dot set state. It gets added to a queue as well. And so when we get to line 21
here, what has actually executed? Well, nothing. And so this dot state dot
count still has a value of 0, and then only after that do
those batch set states run. So what if we actually did care
about the previous count before? Say we wanted to add 1
to the previous count, but we actually really wanted
that state to exist beforehand. Well, we can pass it what’s
called an updater function, and that’s a function that
takes the previous state and returns some new state. In this case, we want it to be the
previous state, dot count plus 1. So now we have two set
states, where rather than just passing an object
to emerge, we’re saying, actually run this function. This function takes the
old state and returns a new object, which is where the count
is the previous state’s count plus 1. And so now when it batches
it, it says, oh, we actually need to run this function twice. And so now when we click,
it actually goes up by two. So any questions on React,
props, or state thus far? Cool. Let’s go ahead and take a quick break. And then when we come back, we can
play with React a little bit more. Hello, and welcome back. So before the break, we were talking
about React and props and state. And now with all three, we can
actually go ahead and start building pretty powerful apps. And so for homework,
for the project zero, you guys have been working
on a to do app, which you’ve been writing in all vanilla JavaScript. And today for the rest
of the class, we’re going to go ahead and together
implement that in all React. So what are some strategies
you may go around doing your thing in vanilla JavaScript? Well, say we had to dos created as list
items, where within each list item, we have input, which has a checkbox. Maybe you’re doing the challenge and
you want to take care of the deletes as well. And then, of course, you’re
going to have some span or some way of displaying text. And so maybe have a couple
different functions. One is to create a to do. One is to delete a to do. And what might you do in those functions
in order to create those to dos? Well, first, maybe you’ll get the text. Then what? Maybe go ahead and using
document dot create element, maybe you want to create a list item. Then what? You’ll probably have
to create the input. Then probably create a button. Create the span. And maybe at the end, you
hook those all up together and append those to the list. And so how might you
take care of delete? Well, find the to do. Maybe delete that. Then what? Make sure to update the counts. And maybe we have to do
that over here as well. So you see how we’re doing this
in a very imperative manner. Using JavaScript, we tell the
browser exactly what we want to do. Well, we know that our to
dos are shaped like this. So first go ahead and get the text. Maybe create that [INAUDIBLE],,
create the checkbox, create all these other things. Append them to each other, and
maybe append them to the list. And so we go ahead, and we created a new
to do, but we had to do a lot of steps in order to get there. And so what might be an
easier way to do this? Well, in to do one, maybe we wanted
to abstract out the creation of the to do itself. So maybe in that new
create to do function, we go ahead and make the list item,
make the input, make the button, make the span. Hook those all up together. But that’s one, little,
discrete part of the UI, right? We’re going to start to
componentize these things. Maybe abstract out a function for
creating the to do, and then in the new to do, we can still get the text, update
the counts, and append to the list. But rather than doing all of this
work in the new to do function, maybe we just invoke create to do. And so we’re starting
to break out something that was pretty complicated into
now two separate steps, where each step is relatively simple. But still in each step, we have
to go about all of these steps in order to get what we want. So what might be a
better way to do this? Well, rather than going through
the steps, where we say, oh, hey, get me a new element called a list
item and input button in a span, maybe we wanted just to write that
all somewhat declaratively ourselves. And so we can actually do
something called innerHTML, so say we created a list item. And maybe we want to take
advantage of JavaScript’s ability to do in-line string
concatenation in order to create some HTML directly here. So we could do something
like li.innerHTML, and we can start actually
putting this directly in here. And so this will actually, in
a more declarative manner, say, hey, I want an input,
a button in the span, and just stick this in the
innerHTML of that list item. And so it’s starting to
look somewhat like something we’ve seen before, right? It’s starting to represent
more of a React component. Do you guys see how that is starting
to look a lot like what we were doing before, using React components? But let’s stick with vanilla JavaScript
still and take this a step further and actually store the to dos in memory. And so what have we been
previously using to store to dos? Well, we’ve just been
adding them to the DOM, and then using the DOM as a way to
remember exactly what our to dos are. But what if we actually wanted to
rather than storing it on the DOM, actually store that in
JavaScript, in some sort of data structure in our JavaScripted memory? Well, that might be easier
to do stuff like delete on. But how are we going
to get that to the DOM? Well, we’re going to actually
have to write a method ourselves. And so maybe we have to dos
stored in memory as an array. Maybe we have something
called a render to do, which will actually
render a single to do. And by render, I mean turn that
to do, an object in memory, into actual HTML, something
that the browser can display. And so that takes care of
rendering a single to do, but how are we going to
render that entire to do list? Well, maybe we want to clear the list
and actually map over the to dos. So we do to dos dot map. What are we mapping? Well, let’s actually render
each of those to dos. And for each of the to
dos that we rendered, go ahead and append that to the list. And so now we went from
storing to dos in memory, creating this concept of rendering,
where we turn a single JavaScript object into an actual thing
that the browser can display. And then we have this
concept called render, which actually renders all of our to dos. And so we go from what used to be
a very imperative way of writing JavaScript, and going from having
a single to do to appending that to the DOM. So now we’re a more declarative way. We’re declaring that a
bunch of to dos may exist. We’re declaring that, hey, here’s
a way to render a single to do and writing a way to turn that
list into a full rendered page. And so now when we want
to add a to do, rather than doing a bunch of DOM
manipulation, maybe we actually want to just do something like a
new to do and push that to the list. Well, once you push it to the list,
now we have to manually say, hey, we updated the list. Let’s go ahead and
render that to the page. And so now if we’re storing
everything in memory, it becomes a lot easier
to remove things. And so how are we going
to remove the to dos? Well, we have to figure out
exactly how to do the to dos, and this is just an
implementation detail. But we can actually reset to
dos to be a filtered list. So just filter them by excluding that to
do that we didn’t want to be included. And then we, again, since
we changed everything, now we’re going to have to go
and manually kick off a render. And so now we’re writing JavaScript in
more of a declarative paradigm, more like what we saw in React earlier. But we’re still implementing stuff like
rendering and rendering a single to do and kicking off these things manually. And let’s actually see how this looks
if we were all to do it in React, where React actually handles a lot
of the data rendering to the screen, using a library rather than stuff
that we have to write manually. So let’s actually do this from scratch. First let me save this. So let’s go ahead and actually, from
scratch, implement our to do list all in React. So of course, we’re going to
have to store stuff in state. So rather than having app be a function,
what might we want to do instead? Have it be a class. And just like in the example
I was showing earlier, we have a concept of rendering that app. React components actually
do this automatically for you in this render method. And so let’s go ahead and
implement that rendering method. And so returning this. And now we have a blank to do list. So now we have a class called app. It’s extending React
component, which gives us what? It allows us to have state,
and it gives us access to this thing called this dot set
state in order to update the state. And this render method goes ahead
and just renders a blank div with a blank UL. Yeah? AUDIENCE: In the previous
version of render, why was it returning a [INAUDIBLE]? JORDAN HAYASHI: Oh. So the question was in my
example to do app three, why was render up here returning false? So we’ll talk about handling
events in a little bit. But basically, browsers actually
all have a way to handle events. And so if you do something
like click a Submit button, it will automatically go ahead
and submit a form for you and refresh the page. And so a lot of times, when we actually
want to handle that all in JavaScript, we have to have a way to disable
the automatic page refresh. And so there are a couple
different ways to do that. You can do something like event
dot stop propagation or event dot– I forget the exact method
name, but it’s something like stop this event from happening. Prevent default. Event
dot prevent default. And a shorthand for doing
both of those is just returning false out of
whatever handler they do. And so when I do return
false from this render, that means if I ever want
to do something, like I add to do, and hook it up to a button
that might automatically submit a form, I can return false from that
instead and automatically prevent that from happening. And so this is just some shorthand
for rather than having submit buttons automatically refresh a
page, it just blocks that and lets me handle it all in JavaScript. Cool. So back to this example in React. So we’re up to this point where
we have a class called app. Right now, we have no state associated
with it, and all we’re doing is rendering this empty list. And so how might we go about
implementing these features that we have in our to do project? Well, first, we’re going to
want to start tracking something in [? our ?] state, right? Maybe we want to track the to dos. And so how do we go ahead and
create something called state? Well, first, we want to
invoke the constructor, which for now takes no arguments, since we’re
not actually passing any props to app. First thing we do in the
constructor is always called the super, which is invoking
the constructor on React.Component, which gives us things
like this dot set state. And then now we have the
opportunity to initialize our state, so we can go ahead and do
this.state equals whatever. And so what might we
want to store in state? AUDIENCE: To do list. JORDAN HAYASHI: Yeah, the to do list. Is there anything else
that we should store? Nothing I can think of
off the top of the head. But thankfully, if we want to add
something later, it’s very easy to do. And so let’s just do to dos, and
let’s store this all on an empty list. Cool. So now we have this concept of state. We now have this concept of to
dos, which is stored in our state, but we’re not doing anything with it. And so let’s actually render
those to dos to the screen. So we might want to do it
within the unordered list. Let’s have something like
this dot state dot to dos. And again, we wrap things
in single curly braces when we want to execute
them as JavaScript. And so we have this dot state dot
to dos, which is an empty array. But we can’t just take for granted
that it’s always an empty array. The point of writing declarative code
is that we want this to hold true, no matter what the state is. And so we just assume that this is
full of to dos, and so we can go ahead and map over that and render each to do. So for each to do, let’s
actually create a to do. All right, so I’m mapping out
the to dos, and for each to do, I’m creating this tag
called uppercase to do. So who remembers what
uppercase tags mean? AUDIENCE: It’s a React component. JORDAN HAYASHI: Yeah,
it’s a React component. And we now have to describe something
called a React component called capital to do. And so let’s go ahead and do that. And so just like the examples
that I showed you earlier, where we abstracted out the
concept of creating a to do, we can do that in React as well. So let’s have this thing called a to do. And what it does is it takes some props. We can call this variable
whatever we want, but the React convention
is to call it props. And what are we going to return? Well, let’s just return an
unordered list, or a list item. What might we have in that list item? Well, we had an input of type checkbox. We had a button that
allowed us to delete. And last but not least,
we had some text. And what was in that text? Well, probably just props, dots– we can call it the text for the to do. Cool. So now we declaratively just said,
hey, every time we want a to do, this is what it looks like. It’s a list item, and
inside it, we have an input. It’s going to be a checkbox. We’re going to have a
button called delete. It doesn’t actually do anything for now. And we also have a span, and
what’s that value of the span? Well, just look at my
props and grab the text. OK, so now what do we have to do? Maybe we should add a way to add to dos. So let’s create a
method called add to do. And what should we do first? Maybe we should get some text. So there are many
different ways to do this. Let’s just for now prompt. And then what do we
want to do with that? We have text. What should we actually do with it? AUDIENCE: Add it to the to do list. JORDAN HAYASHI: Yeah, we should
add it to our to do list. And how are we going to do that? We can just do this dot state
equals something, right? No. There’s only one way
to update state, right? We should do this dot set state. And what are we doing? Well, we have the to do equal
this dot state dot to dos. And then additional thing. And so the way that
we’ve added to an array thus far is by doing
this sum array dot push. What that does is it
actually mutates the array. And so the way that React
knows that it should update is it sees if the values that
are passed to it are different. And as we learned a couple
lectures ago, all objects are actually stored by reference. And so when we go ahead
and mutate an object, the reference doesn’t actually change. And so the React paradigm is
rather than mutating these props, we should actually create new ones. And so a quick way to
create a new array is by using this dot dot
dot notation, which means given an array, if
you dot dot dot array, it just pulls out all of
those values of the array. And so when we do something like
brackets dot dot dot some array, it pulls out all the values of that
array and puts it into a new array. And so this is a quick
way of cloning an array. And so now the array reference is
changed, but the values in there are all the same. And so now we’ve cloned the array,
and now we should add a new to do. And so let’s call our to dos
objects that have a text field, and that can be the text. So there’s a small bug
in our program now. Well, one, we don’t
know how to add a list. Add to that. So let’s go ahead and create
a button that can add a to do. So now you have a button,
but it’s not actually doing anything when we click it. So first we need to set
up an event handler, so when that button is
clicked, what should it do? Well, it should do this dot add to do. So what is wrong with this code? Why isn’t it working? Yeah, this is not bound
to what it should be. And so how are we going to fix that? Well, the easiest way is
to just declare an arrow function in line here, which will
lexically bind what we want it to do. And so now we have another
bug, so we declared to do to be a list item with
an input, a button, and a span. And we see the input in the
button, but we don’t see any spans. So does anybody spot exactly
why that’s happening? So what text are we
populating that span with? Props dot text. And so what does the
object called props– what values are in there? They’re key value pairs that are
mapped to what are passed in here. Are we passing a key
called text into to do? No, we’re not. We’re only passing something
called to do, right? To do may have a value
called text, but in order to actually access that, what we would
have to do is to do props.todo.text. And that will allow us to
pull up the correct field. So now we have a way to add to dos. Great. That doesn’t make much of a to do app. We should be able to check
and uncheck, which we can. But we’re not really
handling that in state. We have no way of telling whether
these to dos are checked or not, and we also have no
way to delete to dos. And so we need some sort of way
to recognize which to do is which. What are some strategies
to go about doing that? Can anybody think of one? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah, we
should give each one an ID. That way, if we can guarantee
that those IDs are unique, then we’ll have a unique way
of recognizing each to do. And so let’s go ahead
and do that real quick. So a very easy way to do that would
be to do something like this– have some value called ID declared
outside, initialized to 0. And every time you create a new
one, you just increment that. And so now every single time we
add a to do, it has an ID field. And we’re guaranteed that
it’s unique, since it’s a counter outside the state of our app. Does that makes sense? So how does that help us? So now we have a way of
recognizing which to do is which. And so say we wanted to get rid of one,
it’s now pretty easy to do that, right? So if we want to now remove to
do, that should now do what? Does it make sense to have a
remove to do with no arguments? Not really, right? Because we need to know exactly
which to do we want to delete. It makes sense that we want to
add a to do with no arguments, because there’s no particular
to do that we want to add. We create that text later. But when we want to
remove a to do, we need to know exactly which to do to remove. And so now we should have remove
to do actually taking argument. And so what argument should it take? Well, we created some sort
of unique identifier field, so we can just take the ID. And so now how are we
going to remove the to do? Well, we could do this.setState, which
is the only way to update the state. And the to dos are what? They can take the old state’s to dos
and just filter them, where for each to do, what do we want to keep? Well, as long as the to do dot ID is
not equal to the ID that we pass in, those are the to dos
that we want to take. Great. But the Delete button is not working. Why not? Well, our component called
to do is not doing anything when you click on the button. So how might we handle that? Well, that’s where props come in. Since we can pass any
JavaScript value as props, what’s to stop us from just
passing a function down? And so when we map over
the to dos like this, we can actually also
pass a function down. So we could pass something called
onDelete and just pass a function. And what would that function do? Well, it’s going to call
this.delete removeTodo. And what are we going to
pass in when we invoke that? Well, this todo.id. Does that make sense? So we’re mapping over to dos,
and for each to do in that array, we’re going to invoke that to
do function, that component. And what are we passing as props? We’re passing it this
thing called onDelete which is a function that invokes
remove to do with that to do’s ID. So for each to do that
we create, we’re passing a unique function that
removes that particular to dos from our global state. And of course, we’re also
passing down the to dos, so it knows exactly
which text to render. And so how do we get that to fire
when we click the delete button? Well, that function’s
just a prop, right? So when we click that
button, what should we do? Well, we should invoke
props dot on delete. And so now when we add things,
if we click that button, the correct one is deleted. Again, how does it know that that first
one is the one that should be deleted? Well, it’s because each
to do that gets rendered, so when we map over the to dos in
our state, we pass a unique on delete handler down. It’s a function that gets bound to
that particular to do’s ID field. And so when we click that
particular to do’s delete button, it invokes remove to do with
that ID, and so it knows to filter based on that particular ID. Does that make sense to people? Great. So we’ve handled the two big parts– well, the add and the delete
part of the assignment. And so what are we missing? We’re missing some way of tracking
how many to dos we have total and how many to dos that we still
have to complete– or in other words, to dos that are unchecked. And so now we have a choice–
how are we going to handle this? Well, if we were back in
Vanilla JavaScript world, a way to do that would
just be to keep track of however many unchecked there are,
keep track of however many total there are. And just mess with those values every
single time we add, delete, or check or uncheck a to do. But in React land, that seems a
little imperative for our like, and so there might be a more
declarative way to do this. But in order to do that, we also
need to track every single to do, whether it’s checked or not. And so currently our to dos are
an ID field and a text field. But maybe it would be better to also add
a field of whether it’s checked or not. And so now when we add to dos, we
add an object with three keys– an ID which is guaranteed to be
unique, because we’re incrementing this ID counter, which lives outside of
our app; the texts that we prompt for; and lastly, a value called checked,
which keeps track of whether that to do is checked or not. And we should just
initialize it with false, because there’s no point
in creating a to do. That’s already done. Cool. And so now we’re still
missing one thing. So when we create to dos, we
just create an empty checkbox. And so no matter what,
it’s going to be unchecked. And so we should set that check field
to equal whether that particular to do is checked or not. And so how might we do that? Well, with props.todo.checked. And so now that keeps track of
whether a to do is checked or not. But the problem is we’re not actually
updating our app state every time a to do is checked or not. And so we’re going to have to
actually handle every single time an input is clicked. Every single time it changes, we need
to update our app state accordingly, because if you look here, I’m
clicking but nothing is happening. Why is nothing happening? Well, we’re initializing our
to dos to be checked if false. We’re saying, for each to do, make sure
that checkbox, whether it’s checked or not, is whether that to do– its check value is true or false. But we have no way to
change that from not false, and so no matter how
many times I click that, it’s not going to change the
value of props.todo.checked. And so we’re going to actually
have to write some logic in order to handle that. And so now let’s create a new
method, called toggle to do. We’re obviously going to need to
take an argument here, because we need to know which to do to toggle. And so what are we going to do here– so again, we’re going to do
this.setState and we can to dos. It’s this.state.todos.map. So if you remember back a few lectures
ago, when we talked about filter, map, and reduce, we talked about how
map applies a function to each of the values in an
array and will return a new array with those
values swapped out with whatever this function returns. And so how might we use that map
function to handle the toggling of a to do? Well, first, which to do
do we actually care about? We only care about the
to do with an ID that matches the ID that were passed, right? So we can immediately
off the bat say, if to do dot ID is not equal to the ID that
were passed, what should we do? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah, we
shouldn’t do anything, right? We should just return that to do. OK, great. We have one condition, but what happens
with the to do that we want replaced? We should flip its checked value, right? And so we’re going to want
to return a new object, and what should those objects’ field be? Well, we can say, ID is to do dot ID. We can do text is to do dot text. And what do we want the
checked value to be? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah, the
opposite of what it was before. And now it looks like we should
have something that works. The problem is it doesn’t. It still doesn’t change anything
when we click this value. Anybody have any ideas why not? It looks like our toggle
to do function should work. We’re setting state,
we’re updating the to dos, we’re mapping over this
with this function, where if the to do’s ID doesn’t
match, we don’t change it. But if it does, we flip
the to check value. But when we click over here,
nothing actually happens. Why might that be? Yeah? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah. Exactly. We defined this function, but we’re
not actually calling it anywhere. And so just like we have this
onDelete that passes a new onDelete function to every single
to do, we should also have this thing called
onCheck or onToggle that passes the unique function
to each to do that we should invoke when we toggle it. And so let’s have an arrow
function here so we don’t have to worry about the this binding. And when on toggle’s invoked, we should
do this.toggleTodo with the todos.id. And so now since we’re
passing a new prop here, we should probably invoke
that prop at some point. And so where are we going to do it? Well, we can do it in the input. We can do something like
onChange=props.onToggle. And how do I know that this
on change attribute exists? Well, you can read the
docs and it’ll be there. Or you can wait till
the next lecture, when we talk about event handling in React. And so now we can create a to do. And when we click it,
look at that– it works. Cool. So now we’re tracking every single
to do, whether it’s checked or not. But that wasn’t really the goal, right? The underlying goal was to keep
track of how many to dos there were, and how many unchecked to dos there are. And so how might we do that? Well, first let’s create some divs. This div says to do count. And then we’re going to have
some JavaScript in line there, and then maybe a div for
unchecked to do count. And again, some JavaScript. Cool. And so what is a quick and easy way
to calculate how many to dos we have? We’re tracking all of
our to dos in memory. What are we using to track them? Yeah? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Yeah. It’s just an array that
has a length attribute, so we can do this dot state dot
to dos, which grabs that list, and just do dot length. And now we have a 0. And if we add a to do, it
goes ahead and increases. Cool. And so how are we going to keep
track of the unchecked to dos? Well, we know we’re tracking every
single to do and whether it’s checked or not. And so one strategy might
be to get all of the to dos, remove all of the ones that are checked,
and then count the number remaining. So what functions might we use
to implement that algorithm? Yeah, we can use the filter function. So we can do this dot state
dot to dos dot filter. And what are we going to filter by? Well, for each to do,
return if it’s not checked. And then we’re left with a new array,
and we can just grab the length. And so if we do a test,
now we have it unchecked. What happens when we check it? Well, that’s going to get filtered out
here, and what’s left is an empty list. And you grab the length
of that, and we have 0. What happens when we delete it? Well, we don’t actually
have to update any logic, because everything is handled
completely declaratively. Is there anything here
that says, every time we change something, make sure to go and
change this count that we’re keeping? Not really, right? We just assume that we have some to
dos and some data that’s coming down, and we just run calculations
based on that data. And so we’re just declaring
what we want, based on any data that’s passed into us. And so React, in that
way, is very declarative. And so this is great. If we want to add any new features, we
don’t really have to worry about, oh, what is every single
possible method that we need to update this new feature with. Whereas if we were doing
this in more Vanilla JS and more an imperative
nature, we would have to make sure, oh, if we want to also
keep track of the count of checks to dos, we’re going to have to
check every single method where that number might be affected
and go ahead and update it there. And so keeping track of
all that in your mind might create bugs that
you might not otherwise if the paradigm were more simple. And so this is a much more simple way
of saying, hey, given some app state, go ahead and render an
app based on that state. Any questions? Yeah? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: Can you
repeat your question? AUDIENCE: [INAUDIBLE] JORDAN HAYASHI: So the
question was if we, rather than replacing the old
state.todos over here with a new array, would it still work, if we
were to mutate the array? And the answer to that is, it depends. It depends on the React
component that you’re using, and it depends on some
other methods that you could write that we’ll talk
about in future lectures. But the fact that it
depends is a bad thing, because then you don’t really know– it depends on what else you’ve
written, whether it will work or not. But if you do it like this, where
it’s a completely new array, it will always work. And we’ll talk about
this in future lectures, but the React paradigm is
always to do things immutably, which means if you ever
change something, replace it with something completely new. And that– be a reference to that. That way, if you’re
ever comparing values, you don’t have any bugs that can
appear there if you’ve mutated a value. So technically it’s changed,
but the reference to that value has not changed. Does that make sense? Cool. Any other questions about
React, props, state, or anything that we’ve seen today? All right. So we talked about to do app.JS
and using React on the web. But why limit React just to the web? So there exists this thing
called React Native, which is a framework that relies on React
Core and the core algorithms implemented by the React library. But it’s actually a framework that
allows you to do much, much more. It allows us to build mobile
apps using only JavaScript. And so when Facebook released this
framework called React Native, they released it with the tagline,
“learn once, write anywhere.” So you only have to
learn JavaScript once. You only have to learn React once. But you can actually write
it, and it’ll run anywhere. And this supports both iOS and Android. And so I’ll leave you there for
this lecture, and next lecture, we’ll look at this thing
called React Native and how to run JavaScript
for mobile apps.

19 comments on “React, Props, State – Lecture 2 – CS50’s Mobile App Development with React Native

  1. this.setState({
    todos: […this.state.todos, {text: text}]
    });

    What this meanings?

    why after …this.state.todos has {text: text}

  2. Excellent material, excellent instructor. Very good at exposing topics! Thank you for uploading this video series!

  3. at 48:40, I dont get it
    "this" under increaseCount() should refer to the App object because increaseCount() is not an arrow function and it is also a method of App object
    So why this.setState() under increaseCount() failed?

Leave a Reply

Your email address will not be published. Required fields are marked *