The adapter supports store.createRecord, store.deleteRecord, and, store.destroyRecord. However, there are some optional configs that you can make use of to support your needs.
The optional configs are available through the adapterOptions property in the save function.
e.g.
import { doc } from 'ember-cloud-firestore-adapter/firebase/firestore';
const newPost = this.store.createRecord('post', { title: 'Post A' });
newPost.save({
adapterOptions: {
isRealtime: true,
include(batch, db) {
batch.set(doc(db, 'users/user_b/feeds/feed_a'), { title: 'Post A' });
}
}
});Indicates if the record will update in realtime after creating it
Type: boolean
Hook for providing a custom collection reference on where you want to save the document to
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| db | Firestore |
Hook for providing additional documents to batch write
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| batch | WriteBatch |
|
| db | Firestore |
The optional configs are available through the adapterOptions property in the save function.
e.g.
import { doc } from 'ember-cloud-firestore-adapter/firebase/firestore';
user.deleteRecord();
user.save({
adapterOptions: {
include(batch, db) {
batch.delete(doc(db, `usernames/${newUser.id}`));
}
}
});Hook for providing a custom collection reference on where the document to be deleted is located
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| db | Firestore |
Hook for providing additional documents to batch write
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| batch | WriteBatch |
|
| db | Firestore |
The optional configs are available through the adapterOptions property.
e.g.
import { doc } from 'ember-cloud-firestore-adapter/firebase/firestore';
user.destroyRecord({
adapterOptions: {
include(batch, db) {
batch.delete(doc(db, `usernames/${newUser.id}`));
}
}
});Hook for providing a custom collection reference on where the document to be deleted is located
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| db | Firestore |
Hook for providing additional documents to batch write
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| batch | WriteBatch |
|
| db | Firestore |
The optional configs are available through the adapterOptions property in the save function.
e.g.
import { doc } from 'ember-cloud-firestore-adapter/firebase/firestore';
post.set('title', 'New Title');
post.save({
adapterOptions: {
isRealtime: true,
include(batch, db) {
batch.update(doc(db, 'users/user_b/feeds/feed_a'), { title: 'New Title' });
}
}
});Indicates if the record will update in realtime after updating it
Type: boolean
Hook for providing a custom collection reference on where the document to be updated is located
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| db | Firestore |
Hook for providing additional documents to batch write
Type: function
Params:
| Name | Type | Description |
|---|---|---|
| batch | WriteBatch |
|
| db | Firestore |
There are 3 types of relationships to take note of: one-to-many, many-to-many, and many-to-none.
When saving a one-to-many relationship, the reference will be persisted on the belongs-to side automatically. For many-to-many and many-to-none however, you will need to manually save them through the include hook on the adapterOptions. This is because:
- Ember Data currently doesn't provide a way to track relationship changes. This means that for the has-many side:
- If a related record exists, we wouldn't know if it was just added or if it was already existing in the first place.
- And, if we remove a related record, we wouldn't be able to know it.
- We want the flexibility of specifying which collection we want the documents to be saved to.
In this scenario, User model has a many-to-many relationship with Group model through the groups and members field name respectively.
import { doc } from 'ember-cloud-firestore-adapter/firebase/firestore';
// Assume that a group1 variable already exists
...
const newUser = this.store.createRecord('user', {
name: 'Foo',
groups: [group1]
});
newUser.save({
adapterOptions: {
include(batch, db) {
// Batch write to the users/<user_id>/groups sub-collection
batch.set(
doc(db, `users/${newUser.get('id')}/groups/${group1.get('id')}`),
{ referenceTo: doc(db, `groups/${group1.get('id')}`) }
);
// Batch write to the groups/<group_id>/members sub-collection
batch.set(
doc(db, `groups/${group1.get('id')}/members/${newUser.get('id')}`),
{ referenceTo: db.collection('users').doc(newUser.get('id')) }
);
}
}
});Saving many-to-none is similar with many-to-many except that you will only batch write to one side of the relationship only.
Another case for many-to-none relationship is to save the record itself in the sub-collection rather than using a reference field. To do so, just simply save the record to the appropriate sub-collection and then push it to the has-many array.
e.g.
In this scenario, the User model has a many-to-none relationship with Reminder model.
import { collection } from 'ember-cloud-firestore-adapter/firebase/firestore';
// Assume that a user1 variable already exists
...
const reminder = this.store.createRecord('reminder', {
title: 'Foo'
});
reminder.save({
adapterOptions: {
buildReference(db) {
return collection(db, `users/${user1.get('id')}/reminders`);
}
}
}).then(() => {
// Update reminders hasMany without flagging user1 as "dirty" or unsaved
user1.hasMany('reminders').push({
type: 'reminder',
id: reminder.get('id')
});
});