I\'m going through the Firestore docs and guide. My code samples below use AngularFire2.
Let\'s consider a \"chats\" collection similar to the examples provided here: ht
The approach you posted seems like it would work and for a large chat application you probably do not want to track every single event that happens in every chatroom as that could be a lot of data. Instead it would probably be better to subscribe to only what is needed and handle periodic updates with cloud functions and cloud messaging.
By using a helper function observeCollection
as well as small code restructuring it would cleanup the service and create observables for each chatroom that would be inactive until they are subscribed to.
class Service {
// db is plan firestore / no angularfire
db: firebase.firestore.Firestore;
loadChatrooms() {
const chatsRef = this.db.collection('chats');
return observeCollection(chatsRef)
.pipe(
map(chats => {
return chats.map(chat => {
return {
chat,
members$: this.observeCollection(chat.ref.collection('members')),
messages$: this.observeCollection(chat.ref.collection('messages')),
};
})
}),
);
}
// Takes a reference and returns an array of documents
// with the id and reference
private observeCollection(ref) {
return Observable.create((observer) => {
const unsubscribeFn = ref.onSnapshot(
snapshot => {
observer.next(snapshot.docs.map(doc => {
const data = doc.data();
return {
...doc.data(),
id: doc.id,
ref: doc.ref
};
}));
},
error => observer.error(error),
);
return unsubscribeFn;
});
}
}
In the application you could then only observe the currently selected chatrooms members and messages, which would save data. Since this post is tagged with Angular async pipes would help with switching by automatically handly subscriptisons.
In your component:
this.currentChat$ = combineLatest(
service.loadChatrooms(),
currentlySelectedRoomId
).pipe(
map(([chats, selectedRoomId]) => {
return chats.first(chat => chat.id === selectedRoomId)
})
);
In your template:
{{ currentChat.name }}
{{ member.name }}
{{ message.content }}