Angular Routing — A Better Pattern For Large Scale Apps

Shai Reznik
Sep 21, 2018 · 6 min read

It’s been a while since my last post because I’ve been busy cooking you people some pretty interesting Angular testing learning material.

But today, I read the following tweet:

And it got me inspired to share my pattern of configuring routes in large scale Angular apps.

Since the beginning, when I read the Angular Routing docs, and saw the “Routing Module” pattern suggestion, I was confused.

“Why do I need a separate module for routing?” I asked.

Nobody answered because I was alone. 🙁

But then I said: “What the hell, let’s try it!” 😎

And I started applying this pattern in one of the Angular 2 (yeah it was a while ago) apps I was beginning from scratch.

Some time went by, and the app grew.

To keep things clean we started separating larger modules into smaller nested modules, but then, I started noticing a HUGE shortcoming when it came down to using nested routes.

Problem #1 — Redundant URL Prefixes

Let’s say we have the following folder structure:



Take details.routing.module.ts for example.

With the “Routing Module” approach we always begin our route URLs from the base URL.

So every route in details.routing.module.ts will need to have the previous routes URLs (“settings/account” ) as its prefix:

@NgModule({ imports: [ RouterModule.forChild([ {
path: 'settings/account/details',
component: DetailsComponent
path: 'settings/account/details/some-other-route',
component: SomeOtherComponent
]) ],
declarations: [DetailsComponent, SomeOtherComponent]
exports: [RouterModule]})export class DetailsRoutingModule{ }

“Why is that a problem Shai?”


for medium to large size apps, these routes will repeat themselves over and over again.

It could slow us down if we ever need to refactor parts of the apps which involve the routes.

We cannot just “plug” a module on top of a different module and expect it to work…

And if you’re thinking “but I can’t just move modules around anyway, it will break my app…”

I got 2 things to say:

  1. Router testing — which I’ll talk about in the future.
  2. Variable based routerLinks — which I’ll talk about later in this post.

Problem #2— Lazy Loading

If for example, we needed to turn details.module.ts into a lazy loaded module, we would have to go ahead and remove all those prefixes from all the details’ routes.

details.routing.module.ts (LAZY)
@NgModule({ imports: [ RouterModule.forChild([
path: 'details', // <-- no more prefix
component: DetailsComponent
path: 'details/some-other-route',
component: SomeOtherComponent
declarations: [DetailsComponent, SomeOtherComponent]
exports: [RouterModule]
export class DetailsRoutingModule{ }

Why? because routes of lazy loaded modules become the children of the parent loading route.


Yeah, you can imagine the child routes of the lazy loaded module “stacking” on top of the parent loading route.

By the way, that’s why we use loadChildren to load lazy loaded modules, as if to say: “Set this sucker’s routes as the loading route’s children”

“Again, why is that a problem Shai?”

In one word: inconsistency.

( In two words: in-consistency 👏 👏 👏 “good job Shai!”)

When we scale up our apps, we want things to be consistent.

We want to reduce the amount of decisions we need to make, so every inconsistency creates unnecessary noise.

“Should I remove the prefix here or leave it? why can’t it just be the same as the other routes…?”

We want to reduce these 👆 questions.

Demo Project To Show The Problem:

Here is an example project that I created for you to see what I’m talking about:

SOLUTION: The “routes.ts” pattern.

To solve these problems I’ve separated the routes array into its own routes.ts file (and removed the routing.module.ts files)




Now, for the eager loaded routes, to achieve the same behavior as the lazy loaded routes (meaning, to stack them up on top their parent route without knowing its prefix) I use this simple trick:

I load the nested routes as the value of the children property under their parent loading route.

export const DETAILS_ROUTES = [ { path: '', component: DetailsComponent }, { path: 'some-other-route', component: SomeOtherComponent },];account.routes.ts
import { DETAILS_ROUTES } from './details/details.routes';export const ACCOUNT_ROUTES = [ { path: '', component: AccountComponent }, { path: 'details', children: DETAILS_ROUTES }];

That way, I keep everything modular and pluggable.

I do this for all the eager loaded routes.

“But what about the lazy loaded module routes?”

Good question!

Let’s say we wanted to turn details.module.ts into a lazy loaded module.

The beautiful thing about this pattern is that it doesn’t require a whole lot of changes to do so.

Instead of loading the DETAILS_ROUTES array in the parent route’s children property, we can just load it in theforChild of the details.module.ts

Like so:

@NgModule({ imports: [
export class DetailsModule { }

And from the account.routes.ts we change the children into loadChildren and load the module itself:

export const ACCOUNT_ROUTES = [ { path: '', component: AccountComponent }, {
path: 'details',
loadChildren: './details/details.module#DetailsModule'

That it!

Everything else stays the same. 🎉 🎉

This way, it’s easy peasy to turn eager modules into lazy ones and still keep the same convention and consistency across your code base.

BONUS: Variables instead of URL strings

I HATE duplicated strings!

Some guy that probably hates duplicated strings as well… who happens to also wear a squirrel suit.

Maybe because I always encounter bugs whenever I use two exact strings in several places throughout my large apps.

So following a widely used pattern from the Java world, I started using variables instead plain strings for route URLs .

That means that alongside each routes.ts file, I now also have a routes.names.ts file which looks like this:

export const accountRoutesNames = {  DETAILS: 'details'}

And I use it both in my routes.ts file :

import { accountRoutesNames } from './account.routes.names';export const ACCOUNT_ROUTES = [{ path: '', component: AccountComponent },{
path: accountRoutesNames.DETAILS,
loadChildren: './details/details.module#DetailsModule'

And in my component files:

import { Component } from '@angular/core';import { accountRoutesNames } from './account.routes.names';@Component({
selector: 'app-account',
template: `
<a routerLink="{{detailsLink}}">Go To Details</a> `
export class AccountComponent { detailsLink = `./${accountRoutesNames.DETAILS}`;}

That way I could refactor with ease, and it won’t ever affect my router links or router tests as long as I keep using variables whenever I need a route URL string.

Here is a project which demonstrates the solution:

To Summarize:

  1. The routing.module pattern has a few downsides: redundant url prefixes and inconsistency in lazy loaded modules routes.
  2. The solution: to use routes.ts files and load them either as children or in forChild in combination with loadChildren
  3. BONUS: use variable names to refer to URLs for better refactoring abilities.

This is the way I’ve been configuring my routes for quite sometime now and I find it to be very consistent and scalable.

I hope this pattern / convention will serve you as well as it has served me and my clients.

Let me know if you have any questions / suggestions in the comments below.

And stay tuned for some exciting news about Angular and Testing… 💪


Follow me on Twitter

Or watch 👇 my Angular courses (especially on Testing):

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store