DEV Community

Cover image for Map vs MergeMap vs SwitchMap

Map vs MergeMap vs SwitchMap

map, mergeMap and switchMap are three principal operators in RxJS that you would end up using quite often. It is necessary to understand what they do and how they differ.

Map

map is the most common operator in Observables. It acts relatively similar to map in Arrays. map takes in every value emitted from the Observable, performs an operation on it and returns an Observable (so the Observable chain can continue).

Imagine it as a function that will take the original values and a projection. The function applies the projection on said values and returns them after transformation.

Let's take an example. Imagine we have an Observable of Array. This Array is a collection of persons. An Object represents each person, and every person has their name and favorite character. We are only interested in getting a list of all characters.

import { of } from 'rxjs';
import { map } from 'rxjs/operators';

const observable = of([
    {
        name: "Parwinder",
        character: "Calcifer"
    },
    {
        name: "Laure",
        character: "Alchemist"
    },
    {
        name: "Eliu",
        character: "X-Men"
    },
    {
        name: "Robert",
        character: "Link"
    }
]);

observable.pipe(
    map(arr => arr.map(person => person.character)) // loops over objects and returns characters
).subscribe(
    char => console.log(char) // ["Calcifer", "Alchemist", "X-Men", "Link"]
);

MergeMap

mergeMap is a combination of Observable merge and map. There are times when your map or projection will generate multiple Observables. For example, now I have an array of characters, and for each character, I would like to make a backend call and get some information.

import { of, from } from 'rxjs';
import { map } from 'rxjs/operators';

const dummyApi = (character) => { // fake api call function
  return of(`API response for character: ${character}`).pipe(
    delay(1000) // the fake api takes 1 second
  );
}

from(["Calcifer", "Alchemist", "X-Men", "Link"]) // characters I need to get information for
.pipe(
  map(arr => dummyApi(arr)) // generates 4 new Observables
).subscribe( // subscribing Observable (outer) of 4 Observables (inner)
  data => data.subscribe(i => console.log(i)) // subscribing to inner Observables
)

Output:

API response for character: Calcifer
API response for character: Alchemist
API response for character: X-Men
API response for character: Link

It works. The output is what we expected. You see the problem here? We are subscribing to what map provides and then subscribing again inside the subscribe block to each Observable supplied by the API call. The approach works, but it is not ideal. This is where mergeMap comes in to play. As I said, it maps, and it merges!

import { of, from } from 'rxjs';
import { mergeMap } from 'rxjs/operators';

const dummyApi = (character) => {
  return of(`API response for character: ${character}`)..pipe(
    delay(1000)
  );
}

from(["Calcifer", "Alchemist", "X-Men", "Link"])
.pipe(
  mergeMap(arr => dummyApi(arr)) // gets 4 Observable as API response and merges them
).subscribe( // we subscribe to one mapped and merged Observable
  data => console.log(data)
)

Pretty neat, huh!

SwitchMap

switchMap does what mergeMap does but with a slight twist. switchMap will subscribe to all the inner Observables inside the outer Observable but it does not merge the inner Observables. It instead switches to the latest Observable and passes that along to the chain.

It still provides one Observable as output, not by merging but by the idea of only emitting the result from the latest Observable.

For our last example, if we use switchMap we will only get the result from the last Observable.

import { of, from } from 'rxjs';
import { switchMap, delay } from 'rxjs/operators';

const dummyApi = (character) => {
  return of(`API response for character: ${character}`).pipe(
    delay(1000)
  );
}

from(["Calcifer", "Alchemist", "X-Men", "Link"])
.pipe(
  switchMap(arr => dummyApi(arr))
).subscribe(
  data => console.log(data) // API response for character: Link
)

There are scenarios where switchMap excels. One such example will be an input box where we provide suggestions to an end-user based on what they have entered (by making an API call for the text in the input field).

If the user is searching for "Chase", they start typing "C", and we make a call. As soon as they type "h", we have to make another call for "Ch". At this moment, our call with value "C" is of no use to us. We should cancel that Observable and subscribe to "Ch" Observable. We need to switch to the latest Observable!

import { of, from } from 'rxjs';
import { switchMap, delay } from 'rxjs/operators';

const dummyApi = (character) => {
  return of(`Search result for keyword: ${character}`).pipe(
    delay(1000)
  );
}

from(["C", "Ch", "Cha", "Chas", "Chase"]) // mimic key input in text field
.pipe(
  switchMap(arr => dummyApi(arr))
).subscribe(
  data => console.log(data) // Search result for keyword: Chase
)

We only get the result for "Chase" Observable, and that is what we want!

Happy coding πŸ‘‹πŸΌ

Oldest comments (2)

Collapse
 
shahab_alam profile image
Shahab Alam Khan

Best!

Collapse
 
emreaka profile image
Emre AKA

Thank you. I don't quite understand the switchMap but I got the others.