guide error handler - devonfw/devon4ng GitHub Wiki
Angular allows us to set up a custom error handler that can be used to control the different errors and them in a correct way. Using a global error handler will avoid mistakes and provide a use friendly interface allowing us to indicate the user what problem is happening.
ErrorHandler
is the class that Angular
uses by default to control the errors. This means that, even if the application doesn’t have a ErrorHandler
it is going to use the one setup by default in Angular
. This can be tested by trying to find a page not existing in any app, instantly Angular
will print the error in the console.
In order to create a custom ErrorHandler
three steps are going to be needed:
In this first step the custom ErrorHandler
class is going to be created inside the folder /app/core/errors/errors-handler.ts
:
import { ErrorHandler, Injectable, Injector } from '@angular/core';
import { HttpErrorResponse } from '@angular/common/http';
@Injectable()
export class ErrorsHandler implements ErrorHandler {
constructor(private injector: Injector) {}
handleError(error: Error | HttpErrorResponse) {
// To do: Use injector to get the necessary services to redirect or
// show a message to the user
const classname = error.constructor.name;
switch ( classname ) {
case 'HttpErrorResponse':
console.error('HttpError:' + error.message);
if (!navigator.onLine) {
console.error('There's no internet connection');
// To do: control here in internet what you wanna do if user has no internet
} else {
console.error('Server Error:' + error.message);
// To do: control here if the server gave an error
}
break;
default:
console.error('Error:' + error.message);
// To do: control here if the client/other things gave an error
}
}
}
This class can be used to control the different type of errors. If wanted, the classname
variable could be used to add more switch cases. This would allow control of more specific situations.
Inside the same folder created in the last step we are going to create the ErrorInterceptor
(errors-handler-interceptor.ts
). This ErrorInterceptor
is going to retry any failed calls to the server to make sure it is not being found before showing the error:
import { HttpInterceptor, HttpRequest, HttpHandler, HttpEvent } from '@angular/common/http';
import { Injectable } from '@angular/core';
import { Observable } from 'rxjs';
import { retry } from 'rxjs/operators';
@Injectable()
export class ErrorsHandlerInterceptor implements HttpInterceptor {
constructor() {}
intercept(req: HttpRequest<any>, next: HttpHandler): Observable<HttpEvent<any>> {
return next.handle(req).pipe(
retryWhen((errors: Observable<any>) => errors.pipe(
delay(500),
take(5),
concatMap((error: any, retryIndex: number) => {
if (++retryIndex === 5) {
throw error;
}
return of(error);
})
))
);
}
}
This custom made interceptor is implementing the HttpInterceptor
and inside the method intercept
using the method pipe
,retryWhen
,delay
,take
and concatMap
from RxJs it is going to do the next things if there is errors:
-
With
delay(500)
do a delay to allow some time in between requests -
With
take(5)
retry five times. -
With
concatMap
if the index thattake()
gives is not 5 it returns the error, else, it throws the error.
Finally, creating a module(errors-handler.module.ts
) is necessary to include the interceptor
and the custom error handler. In this case, the module is going to be created in the same folder as the last two:
import { NgModule, ErrorHandler } from '@angular/core';
import { CommonModule } from '@angular/common';
import { ErrorsHandler } from './errors-handler';
import { HTTP_INTERCEPTORS } from '@angular/common/http';
import { ErrorsHandlerInterceptor } from './errors-handler-interceptor';
@NgModule({
declarations: [], // Declare here component if you want to use routing to error component
imports: [
CommonModule
],
providers: [
{
provide: ErrorHandler,
useClass: ErrorsHandler,
},
{
provide: HTTP_INTERCEPTORS,
useClass: ErrorsHandlerInterceptor,
multi: true,
}
]
})
export class ErrorsHandlerModule { }
This module simply is providing the services that are implemented by our custom classes and then telling angular to use our custom made classes instead of the default ones. After doing this, the module has to be included in the app module app.module.ts
in order to be used.
....
imports: [
ErrorsHandlerModule,
....
As a final step, handling these errors is necessary. There are different ways that can be used to control the errors, here are a few:
-
Creating a custom page and using with
Router
to redirect to a page showing an error. -
Creating a service in the server side or
Backend
to create a log with the error and calling it withHttpClient
. -
Showing a custom made
SnackBar
with the error message.
If the SnackBar is used directly, some errors can occur, this is due to SnackBar
being out of the Angular
zone. In order to use this service properly, NgZone is necessary. The method run()
from NgZone
will allow the service to be inside the Angular Zone
. An example on how to use it:
import { ErrorHandler, Injectable, Injector, NgZone } from '@angular/core';
import { HttpErrorResponse } from '@angular/common/http';
import { MatSnackBar } from '@angular/material';
@Injectable()
export class ErrorsHandler implements ErrorHandler {
constructor(private injector: Injector, private zone: NgZone) {}
handleError(error: Error | HttpErrorResponse) {
// Use injector to get the necessary services to redirect or
const snackBar: MatSnackBar = this.injector.get(MatSnackBar);
const classname = error.constructor.name;
let message: string;
switch ( classname ) {
case 'HttpErrorResponse':
message = !(navigator.onLine) ? 'There is no internet connection' : error.message;
break;
default:
message = error.message;
}
this.zone.run(
() => snackBar.open(message, 'danger', { duration : 4000})
);
}
}
Using Injector
the MatSnackBar
is obtained, then the correct message is obtained inside the switch. Finally, using NgZone
and run()
, we open the SnackBar
passing the message, and the parameters wanted.
You can find a working example of this guide in devon4ts-samples.