I am working with Google Sheets, and I am trying to send data to my Firestore database. I have been able to write to Firestore from Google Sheets, but I can't seem to update a field without completely messing things up.
This is my current testing code:
function getFireStore() {
const email = "your#email.gserviceaccount.com"
const key = "-----BEGIN PRIVATE KEY-----\n your key here \n-----END PRIVATE KEY-----\n";
const id = "project_id";
var firestore = FirestoreApp.getFirestore(email, key, id);
var spreadsheet = SpreadsheetApp.getActive()
var sheet = spreadsheet.getActiveSheet()
var data = {
numIndividuals: sheet.getRange(23, individuals).getValue(),
numTeams: sheet.getRange(23, teams).getValue(),
schoolID: sheet.getRange(23, schoolID).getValue(),
uid: sheet.getRange(23, uid).getValue(),
};
firestore.createDocument("competitions/" + sheet.getRange(23, compId).getValue() + "/registration/abcdefg", data)
}
I understand after playing around with this that it will create a new subcollection titled "registration" with the document "abcdefg." The same thing happens when I use the updateDocument function, as well.
For the website that is reading and writing to this particular Firestore database, I use a similar function .update() to update the document with the correct information. However, in Google Sheets, while it would work the same way it is much more convoluted and tedious to do so.
The way that I came up with for trying to update the document was basically copying everything and adding in the new data.
However, this is seriously tedious and messy. Just copying the data that isn't changed looks like this:
var data = {
compDate: competitions.fields.compDate.stringValue,
contact: competitions.fields.contact.stringValue,
email: competitions.fields.email.stringValue,
grade: competitions.fields.grade.stringValue,
id: competitions.fields.id.integerValue,
maxTeams: competitions.fields.maxTeams.integerValue,
regDate: competitions.fields.regDate.stringValue,
schTeams: competitions.fields.schTeams.integerValue,
schedule: competitions.fields.schedule.stringValue,
site: competitions.fields.site.stringValue,
status: competitions.fields.status.stringValue,
timestamp: competitions.fields.timestamp.integerValue,
user: competitions.fields.user.stringValue,
year: competitions.fields.year.stringValue,
}
The data I want to change is a .mapValue with multiple fields where one of the fields can have multiple fields, which also have multiple fields.
Here's the hierarchy for the field I need to update:
first registration and first team
I know I could do multiple for-loops and whatnot on this, but my question is: is there a simpler way to do this, or do I have to go through and loop over everything to extract only what I want?
As a sidenote, what gets sent to Firestore if I put in the data I got from Firestore using the spread operator, without any editing, it includes every child from the above image. As in, I would have registration -> mapValue -> fields -> 0 -> mapValue -> fields -> etc. And, I don't want those mapValue and fields included, just that actual data (i.e. registration -> 0 -> {schoolID, uid, names, etc.}).
Related
There must be a better way to make upsert in Firebase Firestore in Kotlin.
I have collection of users that contains another collection userDocuments that contains field called highlights containing list of highlights.
I cannot use set and merge options as that will override the highlights list.
Any ideas how to make the code better. I do not like making two database requests on create and handling the failure like this. Maybe my database structure can be also optimized but I thought it is smart as all private userData will be stored in users collections with some subcollections.
My database structure is like this:
users -> {userId} -> userDocuments -> {docId} -> highlights ["this will be highlighted"]
users, and userDocuments are collections. Highlights is a field on userDocument.
docId might not yet be there, there will be 1000 of documents. And I do not want to add it to every user. I want it to be there, only when they make a change such as add or remove highlight to list of highlights.
usersCollection
.document(userId)
.collection("userDocuments")
.document(docId)
.update("highlights", FieldValue.arrayUnion(text))
.addOnFailureListener { err ->
// TODO should be handled differently
if (err is FirebaseFirestoreException &&
err.code === FirebaseFirestoreException.Code.NOT_FOUND
) {
val highlights = listOf(text)
usersCollection
.document(it)
.collection("userDocuments")
.document(docId)
.set(mapOf("highlights" to highlights), SetOptions.merge())
}
}
You can update using dictionary notation or dot notation too.
https://firebase.google.com/docs/firestore/manage-data/add-data#update_fields_in_nested_objects
db.collection("userDocuments")
.document(docId)
.update({
"highlights": FieldValue.arrayUnion(text)
});
You can consider using transactions as I mentioned in the comment above. But not sure if that is what you are looking for.
I'm working on the Flutter app where users can save multiple addresses. Previously I used a real-time database and it was easier for me to push data in any child with a unique Id but for some reason, I changed to Firestore and the same thing want to achieve with firestore. So, I generated UUID to create unique ID to append to user_address
This is how I want
and user_address looks like this
And this is how it's getting saved in firestore
So my question Is how I append data with unique id do I have to create a collection inside users field or the above is possible?
Below is my code I tried to set and update even user FieldValue.arrayUnion(userServiceAddress) but not getting the desired result
var uuid = Uuid();
var fireStoreUserRef =
await FirebaseFirestore.instance.collection('users').doc(id);
Map locationMap = {
'latitude': myPosition.latitude,
'longitude': myPosition.longitude,
};
var userServiceAddress = <String, dynamic>{
uuid.v4(): {
'complete_address': completedAddressController.text,
'floor_option': floorController.text,
'how_to_reach': howtoreachController.text,
'location_type': locationTag,
'saved_date': DateTime.now().toString(),
'user_geo_location': locationMap,
'placeId': addressId
}
};
await fireStoreUserRef.update({'user_address': userServiceAddress});
If I use set and update then whole data is replaced with new value it's not appending, so creating a collection is the only solution here and If I create a collection then is there any issue I'll face?
You won't have any issues per se by storing addresses in a separate collection with a one-to-many relationship, but depending on your usage, you may see much higher read/write requests with this approach. This can make exceeding your budget far more likely.
Fortunately, Firestore allows updating fields in nested objects via dot notation. Try this:
var userServiceAddress = {
'complete_address': completedAddressController.text,
'floor_option': floorController.text,
'how_to_reach': howtoreachController.text,
'location_type': locationTag,
'saved_date': DateTime.now().toString(),
'user_geo_location': locationMap,
'placeId': addressId
};
await fireStoreUserRef.update({'user_address.${uuid.v4()}': userServiceAddress});
I want to add 2 collections in Firestore in React Native.
Like JOIN can be used to add 2 tables. Is there any alternative for JOIN in Firestore to add collections?
I want to add these 2 collections users and users_2
How can I do this? Please help
At the time of writing it is not possible to query documents across collections in Firestore (it is apparently a feature that is on the roadmap however, see this recent blog post https://cloud.google.com/blog/products/databases/announcing-cloud-firestore-general-availability-and-updates -see bullet point "More features coming soon"-).
So that means that you'll have to issue two queries (one for each table, to get all the collection docs) and join/combine their results in your front end.
Another approach would be to duplicate your data (which is quite common in NoSQL world) and create a third collection that contains copies of all the documents.
For this last approach you could use a Batched Write as follows (in Javascript):
// Get a new write batch
var batch = db.batch();
var docData = {email: 'test#gmail.com', fullname: 'John Doe'}
// Set the value of doc in users collection
var usersRef = db.collection('users').doc();
batch.set(usersRef, docData);
// Set the value of doc in the allUsers collection (i.e. the third collection)
var allUsersRef = db.collection('allUsers').doc();
batch.set(allUsersRef, docData);
// Commit the batch
return batch.commit().then(function () {
// ...
});
I am trying to access to a object into another object in my firebase database, i have a structure like this:
I want to get all the objects that have the email that i send by parameters, i am using .child to access to the childs into my object but i am not success with the query, this is my code
$ ref_db.child("/groups").child("members").orderByChild("email").equalTo(email).once("value", (snapshot)=>{
console.log(snapshot.val());
});
The snapshot.val() always is undefined.
could you help me with the query?
One efficient way to get "all the groups that have that email inside the members object" would be to denormalize you data and have another "main node" in your database where you store all "members" (i.e. their email) and the "groups" they belong to.
This means that each time you add a "member" node under a "group" (including its email) you will also add the group as a child of the member email, in this other "main node".
More concretely, here is how would be the database structure:
Your current structure:
- groups
- -LB9o....
...
- members
- -LB9qbd....
-email: xxxx#zzz.com
- -LBA7R....
-email: yyyyy#aaaa.com
And the extra structure:
- groupsByMembers
- xxxxxx#zzzcom
- Grupo1: true
- yyyyy#aaaacom
- Grupo1: true
- Grupo2: true
- bbbcccc#dddcom
- Grupo6: true
- Grupo8: true
Note that in the "extra structure" the dots within an email address are removed, since you cannot include a point in a node id. You will have to remove them accordingly when writing and querying.
This way you can easily query for the list of groups a member is belonging to, as shown below. Without the need to loop several times over several items. This dernomalization technique is quite classic in NoSQL databases.
const mailToSearchFor = xxxx.xx#zzz.com;
const ref = database.ref('/groupsByMembers/' + mailToSearchFor.replace(/\./g, ''));
ref.once('value', snapshot => {
const val = snapshot.val();
for (let key in val) {
if (val.hasOwnProperty(key)) {
console.log(key);
}
}
});
In order to write to the two database nodes simultaneously, use the update method as explained here https://firebase.google.com/docs/database/web/read-and-write#update_specific_fields
This is because you have a random key before members, you need to go through the path and not skip a node, to be able to access the values:
ref_db.child("groups").child("-LB9oWcnE0wXx8PbH4D").child("members").orderByChild("email").equalTo(email).once("value", (snapshot)=>{
console.log(snapshot.val());
});
Below is the piece of subscription on UI.
Template.AssociateEmp.onCreated(function(){
this.validEmail = new ReactiveVar('');
const tpl = this;
this.autorun(() => {
var email = this.validEmail.get();
this.subscribe('GetUnassociatedUser', email, {
onReady: function () {},
onError: function () {}
});
});
});
Is there a way to know that even if the dynamic data changed (here validEmail), Meteor Subscription was unaffected and did not change its data on UI? Is there any flag or something that triggers when subscription data is unchanged?
Autorun, ReactiveVar and Subscriptions
In your code example the subscription itself will re-run the server's publication function as the subscription's input variable email depends on the reactive variable validEmail and thus triggers the autorun when validEmail changes.
You can easily check that on your server console by logging something to the console within the publication.
If validEmail remains unchanged than there is no reason for autorun to trigger (unless there are other reactive sources that may not be added to your code example).
What about the subscribed data
Now if something has caused the subscription to re-run and you want to know if the data of a collection has been changed you could easily check on collection.count() but this could be flawed.
Imagine your publication is parameterized to include different fields by different parameters, then the data that is transfered to the client side collection will be different.
You would then require a method to check on the client side collection's data integrity.
Use hashes to verify integrity
A possible help would be to generate hases from the dataset using the sha package.
You could for example create one hash of your whole collection:
// get data
const data = Collection.find().fetch();
// map data to strings
// and reduce to one string
const hashInput = data.map(doc => JSON.stringify(doc) ).reduce((a, b) => a + b);
// generate hash
const collectionHash = SHA256(hashInput);
After the next onReady you can generate a new hash of the collection and compare it with the previous hash. If they are different, then something has changed.
This also removes the need for iterating the collection's documents if you only want to know if the data has changed but it won't reveal which document has changed.
Hashing single documents
Hashing single documents gives you more insight about what has changed. To do that you only need to create a map of hashes of your collection:
// get data
const data = Collection.find().fetch();
// map data to strings
const hashes = data.map(doc => { _id: doc._id, hash: SHA256( JSON.stringify(doc) ) });
You can store these hashes together with a document's _id. If the hash of a document is different after a new subscription you can assume that the change is related to this document.
General notes
hashing is some kind of expensive operation so it might be difficult to keep up with performance on large collections
usually you should design your pub / sub and autorun in a way that when the input changes the output changes
code is cold-written so it may not work out of the box. Please let me know if something does not.