Advanced Routing Strategies

[00:00:00]
>> Scott Moss: That was just one level. This example is showing you how you can go multiple levels, so I can mount another router on the router that was mounted. I can, I can go infinitely deep. There's really no nothing stopping you from doing that.

[00:00:11]
There's really no nothing stopping you from doing that. This is really good if you wanna like segment your router, your API based off of like different versions You're using API was it's like slash V1, you know, V2 like this is great for that. You can segment all that stuff off or maybe you segment them based off of. Microservices and different teams work on different Microservices so everybody gets their own sub router and that's why you do it that way because it's an organizational thing maybe it's a technical thing it could be any reason in which you might want to separate these things out.

[00:00:27]
Microservices and different teams work on different Microservices so everybody gets their own sub router and that's why you do it that way because it's an organizational thing maybe it's a technical thing it could be any reason in which you might want to separate these things out. You can go infinitely deep if you want to. There really are no it I would say the deeper you Go, just realize at some point all these things coalesce into one registry of all your routes, so you have to be wary of possible collisions that we had that I that I discussed before about like which one's gonna run first. If you keep going deep and there's a potential collision with some other nested andset or, you know, great second cousin route over here that you don't know about and there's a potential conflict, which I only got registered first, that's the one that's gonna happen.

[00:00:44]
If you keep going deep and there's a potential collision with some other nested andset or, you know, great second cousin route over here that you don't know about and there's a potential conflict, which I only got registered first, that's the one that's gonna happen. So if you have a sprawling router configuration that is like that could be nasty to debug, but yeah, in that case, I think that's where you will. Move on to better hygiene with registering routes by like maybe instead of just writing strings like this, you would have like enums and globals for your route names so therefore you can reuse them and things like that versus just like somebody just type in whatever stream they want so I dive into a little more about just because it's just JavaScript and there's nothing stopping you do whatever you want, you can conditionally make routes, right? So like, hey, if I'm in this environment, if I'm in dev mode.

[00:01:04]
So like, hey, if I'm in this environment, if I'm in dev mode. I wanna, I wanna add these Dv routes specifically for dev mode, you know, maybe there's like something you have on your Web app. That's like a deaf panel that shows certain stuff and it needs to get access to a specific thing in the database so you only enable these routes if you're in dev mode so you do that, right? Or maybe there's like feature flags that are powered by environment variables that you set inside your hosting provider and that feature only gets enabled if that environment variable is set to true so then you add the route for that feature, right?

[00:01:26]
Or maybe there's like feature flags that are powered by environment variables that you set inside your hosting provider and that feature only gets enabled if that environment variable is set to true so then you add the route for that feature, right? Same thing like AB test or things like that. He here's the versioning thing that I was talking about so you can have different versions of different stuff. So there's many reasons why you might wanna do that.

[00:01:43]
So there's many reasons why you might wanna do that. Also just talk about like how to organize your routes and in this app we're gonna organize them. There's a lot of ways to do it, but we're mostly gonna organize them by like resource based, so. What does that mean is like every resource, like a habit or a user or you know off in this case, which is technically not a resource we're gonna separate each thing out by the resource by that there's really no wrong way because I've also done this other approach where it's like feature based so you could think of features as in like what someone is going to experience from a user's perspective and you can break everything out to those separate features.

[00:01:56]
What does that mean is like every resource, like a habit or a user or you know off in this case, which is technically not a resource we're gonna separate each thing out by the resource by that there's really no wrong way because I've also done this other approach where it's like feature based so you could think of features as in like what someone is going to experience from a user's perspective and you can break everything out to those separate features. This might make sense if different teams are working on different features, stuff like that. You might also break it down by like different versions, you might also break it down by like. Like you might co-locate things right?

[00:02:13]
Like you might co-locate things right? Like, for instance, on this resource based one, we have, we have a routes folder and then we have all the resources, but maybe instead of having a routes folder you might have a off folder and a user folder and a habit folder and a tag folder, and each one of those folders has the routes for that resource the test for that resource The handlers for that resource the database queries for that resource, everything for that one resource is in that one folder and then you find a way to combine them all collectively into some index file somewhere, right? So there really is no wrong way. These are just different examples that I personally have used in the past for different reasons, but if it was just me and I'm just like getting stuff out the door, I'm just gonna do this cause it's simple, it's flat.

[00:02:28]
These are just different examples that I personally have used in the past for different reasons, but if it was just me and I'm just like getting stuff out the door, I'm just gonna do this cause it's simple, it's flat. Again, I talk about testing. This is just to give you an example of that. If you wanted to test some of the things we just did, we will be writing tests later so you don't need to worry about that, And then I added this section here for just like common endpoints, so I talked about the health one initially, a lot of servers have slash health in which there's tons of apps that you can go pay for right now that will literally just ping your like you been to like a.

[00:02:46]
If you wanted to test some of the things we just did, we will be writing tests later so you don't need to worry about that, And then I added this section here for just like common endpoints, so I talked about the health one initially, a lot of servers have slash health in which there's tons of apps that you can go pay for right now that will literally just ping your like you been to like a. What is it? Is it, is it just ping? There there's something called ping, but I guess Uptime robot might be more, OK, yeah, here what I'm looking for.

[00:03:06]
There there's something called ping, but I guess Uptime robot might be more, OK, yeah, here what I'm looking for. I'm looking for this view right here. We've all seen like a graphic like this where it's like, here's the name of the service and there's like this bar, and then like if there was like a drop in the service, maybe this bar would be like slightly red or yellow or something like that. OK, part of how they're doing that is they're just pinging that server over and over and over again.

[00:03:20]
OK, part of how they're doing that is they're just pinging that server over and over and over again. And as long as they get a 200, everything's good. As soon as they don't get a 200, they log that at that time and that shows up in this graph. So and people get alerts and stuff like that.

[00:03:36]
So and people get alerts and stuff like that. There's different thresholds like, oh if we get. If we get, you know, X amount of 500s in the span of X amount of minutes then ping this person on call or something like that, right? So.

[00:03:48]
So. Pinging is a very effective way. Of determining not only is your service still up, but how long is it taking? Because if you keep pinging your website and it increasingly takes longer that means.

[00:04:07]
Because if you keep pinging your website and it increasingly takes longer that means. Something's not right You need to figure out what that is. Is it the traffic that you're getting and congrats, you found product market fit, or is it that code that so and so just pushed up and like killed the whole app? Like what is it?

[00:04:23]
Like what is it? So. Health checks paying is a very well known thing that people do. You do like a more detailed one if you wanted to.

[00:04:40]
You do like a more detailed one if you wanted to. API versioning I talked about, as well. There's also like catch all routes. I don't ever do this in practice.

[00:04:55]
I don't ever do this in practice. It's why I'm not really showing you, but I thought it'd be a crime. Well, that's why I'm not really putting it in the app, but I'm showing you because I thought it'd be a crime not for you to know how to do it. But yeah, you could do catch all routes which is simply like.

[00:05:09]
But yeah, you could do catch all routes which is simply like. You know, In this case it's like if anybody goes to slash API anything automatically 404 you might be thinking why would I ever do that? Well, remember order matters so if you put this at the bottom of all the API routes that you already declared that means any other route that isn't already declared that has API in it will then go here So you're overriding the default 404 that Express would have gave anyway with your own custom 404 for any route that has API in it that wasn't previously registered. Same here, like you could say, hey, if anybody.

[00:05:23]
Same here, like you could say, hey, if anybody. Oh man, hell yeah, I forgot I remember right, OK. Back in the day when you made a SPA app single page application with client routers like React router and all that stuff, you still have to do this by the way, you just don't do it anymore manually, but there had to be a server that when hit will always send back the index to HTML no matter what. That actually still has to happen today for a single page app.

[00:05:36]
That actually still has to happen today for a single page app. This is why your single page apps with routes probably won't work. Without some type of server sending back a file and that's because that's why it's called a single page app. You only ever send back the index the HTML file and from there the JavaScript loads up and the browserrouter, the HTML 5 push date router kicks off and it takes over Routing and it remembers everything, so you have to make sure that when someone tries to go to.

[00:05:52]
You only ever send back the index the HTML file and from there the JavaScript loads up and the browserrouter, the HTML 5 push date router kicks off and it takes over Routing and it remembers everything, so you have to make sure that when someone tries to go to. Some route on your site and which is going to for sure initiate a server request that the server just responds back with the index HTL like look. Ask to ask to ask React the same me like I'm just gonna send back the index HTML. So you have to do this manually back in the day.

[00:06:04]
So you have to do this manually back in the day. Now you actually get this for free with any dev server that you've ever created like if you just use like Vite Dev server or anything they're doing this for you automatically assuming you're not doing service side rendering, which is a whole another thing. So, yeah, and if you deploy like a single page application on like a What is it? What do they call them now, like a jam stack deployable thing, like a netfly or something.

[00:06:21]
What do they call them now, like a jam stack deployable thing, like a netfly or something. This is what they're doing, they're doing this for you, so. Talk about orders, parameters. Just some common pitfalls here.

[00:06:41]
Just some common pitfalls here. People, you know, when I say people, I mean me. Sometimes I forget to export the router, be sure to do that. route order We'll get you messed up.

[00:07:00]
route order We'll get you messed up. Oh, this is, this is a common one. I don't know why I forgot to actually show this one. This is so common.

[00:07:17]
This is so common. These are, these two are actually the exact same route. Do, does that make sense? Why these two would be the same route?

[00:07:39]
Why these two would be the same route? Right, cause think about it, slash ID means slash anything. And then this is slash profile, which is technically slash anything. So these two are the exact same route.

[00:07:57]
So these two are the exact same route. So that means if slash ID was first. As a GET request before slash profile, you would never hit this route. Cause if you did a GET request to slash literally anything, it will always go here.

[00:08:13]
Cause if you did a GET request to slash literally anything, it will always go here. It will never go here. So you have to put this one first. Does that make sense?

[00:08:29]
Does that make sense? It's those things. I think I still run into that today. I might have ran into it in the code that I just wrote.

[00:08:44]
I might have ran into it in the code that I just wrote. I don't even know, but it's just, it happens. Missing route Handlers, you'll get an error from Express if you forget to put a route handler here. It'll say like, oh, I expected a thing, but that happens to me a lot.

[00:09:03]
It'll say like, oh, I expected a thing, but that happens to me a lot. I don't worry about Ayn handlers, and yeah,